Interpreting IDensification  Harris Constantinou

advertisement
UCLWPL 2011.01
81
Interpreting IDensification
Harris Constantinou
Abstract
We put forward a new theory aiming to account for the interpretation of the English intensifier (i.e.
himself as in John himself met Mary). We show that its interpretation is regulated by the interplay
between information structure, semantics and surrounding context. More concretely, we propose (a)
that the intensifier can assume any role of the component of information structure, which we assume to
consist of the autonomous notions of [contrast], [topic] and [focus] (Neeleman and van de Koot 2008,
Neeleman and Vermeulen 2010); (b) that the intensifier always denotes an identity function (ID)
which takes a nominal constituent x as its argument and maps it onto itself (Eckardt 2001, Hole 2002,
Gast 2006). and (c) that specific contextual (both sentential and pragmatic) factors can influence its
interpretation. The theory assumes only one lexical entry for the intensifier and derives its properties
through independently motivated assumptions, thus claiming superiority at least over polysemous
analyses (i.e. Edmondson and Plank 1978, König 1991, Siemund 2000, Eckardt 2001). This proposal
also undermines accounts that assume one lexical entry but derive each use of the intensifier from its
structural position (i.e. Gast 2006).
Keywords: intensifier, contrast, topic, focus
1. Introduction
Examples (1) - (3) constitute a minimal triplet differing only with regard to the position of the
intensifier. In (1) himself is found immediately next to a nominal constituent and this sentence is
roughly understood as the king (in person), and not someone else, performs the action described
by the predicate. We will be referring to this instance of the intensifier as the in person reading.
In (2) himself is found immediately after the auxiliary verb and the example can be paraphrased
with also or too (König and Siemund 1999) (henceforth called the also reading). In (3) himself is
found in a sentence final position, giving rise to a paraphrase containing alone or without help
(König and Siemund 1999) (henceforth called the alone reading). Note that these approximate
paraphrases are merely used throughout most of this paper for the purposes of exposition and avoidance
of any confusion between the many readings of the intensifier. They have no theoretical significance.
(1)
(2)
(3)
The king himself has come to the meeting (and not his secretary).
(Apart from the king‘s secretary,) The king has himself come to the meeting.
The king has come to the meeting himself (without anyone accompanying him).
As shown in the examples above, the intensifier is never found in an A-position and hence does
not receive a θ-role. Therefore, syntactically it is a modifier (or adjunct). Furthermore, it is
always understood to interact with a nominal constituent, even if they are not found next to each
other. In the case of (1) - (3), this nominal constituent is the subject with which the intensifier

I gratefully acknowledge the Leventis foundation and AHRC (UK) for their valuable financial support. I
would also like to thank my supervisor, Dr. Hans van de Koot for his advice and numerous comments on earlier
drafts of this paper.
UCLWPL 2011.01
82
agrees in person, number and gender1. Despite the apparent interpretational dissimilarity of the
examples above, which has led to analyses based on diametrically opposed assumptions, most
researchers agree on one point: an intensifier evokes alternatives to the referent of the nominal
constituent it interacts with and compares this referent with these alternatives. Combined with
the fact that the intensifier is universally stressed and carries some sort of accentual prominence
within the sentence, many analyses (Eckardt 2001, Hole 2002, Gast 2006, among others) suggest
that the inducing of alternatives arises from its interaction with the focus structure of the
sentence (Rooth 1985, 1992).
We think that the above view is on the right track. However, we believe that it has not been
explored in sufficient detail, precisely because previous researchers have underestimated the role
of information structure in their attempts to explain the phenomenon under discussion. This
paper advocates that the properties of the intensifier can be further elucidated if we accord a
more central role to information structure. The main thesis to be defended is that the intensifier is
effectively an information structural device, in the sense that its interpretation is governed by the
interpretative contribution of contrast, topic or focus, or a possible combination of them2. These
notions comprise a particular component of information structure as described by Neeleman and
Vermeulen (2010)3. This hypothesis is flexible enough to derive all the readings that have been
observed. For example, we propose that the alone use of the intensifier is an instance of
contrastive focus whereas the so-called as for use (see example (15) is an instance of contrastive
topic. Their interpretational difference is primarily a result of the semantics associated with these
two information structural notions.
1
This type of φ-feature specification is by no means universal. For instance, the Dutch intensifier is not
specified with any φ-features. Thus the sentence in (a) is ambiguous between the two readings in (b-c).
a.
Jan
heeft
Marie
zelf
een kado
gegeven.
John
has
Mary
x-self
a present
given.
b.
John
himself
has
given
Mary
a present.
c.
John
has
given
Mary
herself
a present.
2
This generalization becomes clearer later in this paper, when we discuss how the intensifier is semantically
specified. In a nutshell, we argue that its lexical entry is specified with ID, a truth-conditionally meaningless
function that maps its satisfier onto itself. Naturally, this exceptionally basic semantic specification leaves room for
an information-structure oriented approach.
3
We are not the first to follow this line of reasoning. In the same spirit, Féry (2010) proposes that the German
intensifier selbst can be an instance of a free focus. However, her account remains largely inconclusive with regard
to the relatively free interpretation of the intensifier. Also, her account diverges from ours with respect to the reasons
found behind the uses of intensifiers. She claims that the different interpretations are a result of the different
domains of the free focus intensifier. These domains are determined by the structural position of the intensifier. By
contrast, we argue that the intensifier’s interpretation is not determined by its structural position and that its different
uses are primarily a result of their characterisation as a contrastive topic, contrastive focus, topic or focus. Bergeton
(2004) also pursues a line of research similar to Féry’s by suggesting a focus-based analysis. He distinguishes four
readings of the intensifier, namely the three shown in (1) - (3) plus a reading paraphrasable to even. Apart from the
fact that he does not account for all the readings that the intensifier can exhibit, he assumes a different approach to
derive the readings, one that is too strict, we believe, to account for all available interpretations. Due to lack of
space, we do not outline Bergeton’s approach.
UCLWPL 2011.01
83
Section 2 surveys some previous analyses and paves the way for our proposal. Our purpose
is not to offer a full review of these accounts, or by any means of the abundant literature around
the topic, but simply to expose some of the fundamental problems of certain assumptions that
persist throughout the literature and introduce some of the notions that will be required for our
analysis. Section 3 develops the core proposal of this paper, and derives the various
interpretations of the intensifier. Section 4 concludes the discussion.
2. Some remarks on previous accounts
2.1. Previous accounts
On the face of it, examples such as (1) - (3) suggest that one could try to link the interpretation of
an intensifier to its position in the sentence. According to prior literature on the topic there are
two ways in which one could go about doing this. One possibility is to assume that every
interpretation of the intensifier is listed separately in the lexicon and that each lexical realisation
is compatible with only a certain position within the sentence. Alternatively, one could assume
that there is only a single lexical entry with a fixed base-generated position. Its various surface
positions would then be derived through movement. On this second approach, the difference in
interpretation associated with each position is attributed to a variety of contextual (syntactic,
semantic, or pragmatic) reasons4. Edmondson and Plank (1978) follow the former line of
reasoning and distinguish intensifiers in terms of their position. In their terminology, himself1 is
always attached to the nominal phrase it modifies (as in (1)). The function of himself1 is to place
the referent of the nominal it modifies in the highest position on a scale of remarkability. In other
words, this instance of the intensifier marks the referent it interacts with as the least expected
person/thing in the situation described by the sentence. On the other hand, himself2 is never found
attached to the nominal it interacts with but still orders its referent with regard to a scale. In this
case, however, the scale is defined in terms of direct involvement by characterising the referent
of the intensified DP as the value most directly involved in the situation. As Cohen (1999) points
out, even though the notions of remarkability and direct involvement are suggested to constitute
a central component of the meaning of himself1 and himself2 respectively, they do not shed any
light on the intensifier‘s function. Examples (4) - (5), taken from Cohen (1999), show that these
notions are irrelevant to the semantic contribution of the intensifier.
(4)
The staircase wound round the lift shaft and went the whole way to the roof, though the lift
itself went no further than the top floor. (Sayers 1986: p.51)
My grandmother knows these things. She is a witch herself.
(5)
4
As mentioned in the introduction, we do not intend to offer a complete literature review of previous
proposals, but simply to briefly consider a few representatives of each line of reasoning. This is both for reasons of
space and because the arguments presented here apply to all proposals which assume an absolute correlation
between the position and interpretation of the intensifier, whether through a polysemous analysis (Moravcsik 1972,
Dirven 1973, Edmondson and Plank 1978, König 1991, König and Siemund 1999, Siemund 2000,) or a
monosemous one (see Gast 2006 for a movement approach, Primus 1992 for an analysis attributing the various
interpretations to the nature of the constituent the intensifier attaches to). For an extensive literature review and
criticism of the proposals outlined here see Cohen (1999), Eckardt (2001), Siemund (2000), Gast (2006), Tavano
(2006).
UCLWPL 2011.01
84
In (4), the lift, despite being modified by himself1, is not understood as being ordered in a scale
of remarkability or even as being remarkable at all. A similar problem is raised by (5), an
instance of himself2, in which my grandmother is not understood as being more directly involved
in being characterised as a witch compared to the rest of the people who are witches. König
(1991, 2001), König and Siemund (1999) and Siemund (2000)5 also follow the first line of
reasoning and (at least) draw a clear link between focus particles such as also, only, and too and
the intensifier. They distinguish three different intensifiers, namely the adnominal (in person
reading), the adverbial inclusive (also reading), and the adverbial exclusive (alone reading).
What seems to be the underlying connection between these three lexical entries is the notion of
centrality (see Baker 1995 for a similar notion, namely of discourse prominence). According to
König and Siemund, all intensifiers characterise the referent of the DP they interact with as
‗central‘, and oppose it to a set of alternative values which are ‗peripheral‘ to the head DP. For
example, the king in (1) - (3) is not compared with just any alternative but only with ones that are
peripheral to the king, such as the king‘s family or the king‘s staff. Apart from their
interpretational variation (as described in the first paragraph of the present paper), what
distinguishes the adverbial types from the adnominal is their distribution. As its name indicates,
the adnominal intensifier is assumed to be a nominal modifier, hence is usually found next to an
NP, whereas the adverbial intensifiers are assumed to be verbal modifiers, hence occurring
somewhere in the VP domain. The main argument, however, in favour of the view that the three
uses of the intensifier are a clear case of polysemy is the ability to build grammatical (laboratory)
sentences exhibiting all the uses simultaneously, as the example below of Siemund (2000)
shows.
(6)
Bill himself has himself not found the answer himself.
However, Gast (2006) shows that the above example does not necessarily illustrate the
polysemous nature of the intensifier. He follows König and Siemund with regard to the possible
interpretations that the intensifier can have (i.e. exclusive adverbial, inclusive adverbial,
adnominal), assuming though a single lexical entry, whose denotation is the identity function ID
that takes a nominal constituent x as its argument and maps it onto itself. This minimal semantics
combined with the assumption that the intensifier always interacts with the focus structure of the
sentence derives some of its interpretational properties (such as the evoking of alternatives). In
particular, he assumes that the intensifier is always base-generated in a position attached to the
DP it modifies. Its various positions are accounted for by assuming that the complex of DP and
associated intensifier undergoes several steps of movement and may strand the intensifier at any
point in the course of the derivation. This is illustrated in (7) for the exclusive interpretation.
(7)
I1 always [TP t1 [do my homework]2 [vP [DP t1 myself] t2 ]].
In this example the following operations are assumed to take place: i) do is moved to T0, ii) the
complement my homework is pied-piped to satisfy the phonological constraint FINALFOCUS
(see Larson 1988 for an analysis of pied-piping that fits this proposal and Büring 2001 for
discussion of FINALFOCUS) iii) the subject DP associated with the intensifier, I, initially moves
5
For an extended criticism of this approach see the works themselves in which the authors recognise that their
analysis faces some serious problems mostly related to the parallelism with focus particles (König 1991, König and
Siemund 1999, Siemund 2000, König 2001) or Eckardt (2001), Bergeton (2004), Gast (2006).
UCLWPL 2011.01
85
to [Spec,TP] and then to [Spec,FinP]6. As is evident from the derivation, the movement of
material across the intensifier is licensed by the FINALFOCUS constraint (the subsequent
movement operation is assumed to result from [Spec,FinP] attracting the subject). However,
when it comes to the inclusive intensifier, which, according to Gast, can occur either in postauxiliary position or sentence-finally, the same phonological constraint is suspiciously less
stringent7. Regarding the interpretational differences between the adnominal and the two
adverbial instances of the intensifier, Gast suggests that they result from their structural position,
which in turn forces them to interact with different types of constituents. Whereas the adnominal
only interacts with a DP, the two adverbials interact with the entire predication. The
interpretational difference between the two adverbial instances is the outcome of the assumption
that the inclusive use is structurally located outside the VP, whereas the exclusive use is located
within the VP. This assumption is based on the argument that the exclusive use is invariably
within the scope of negation, whereas the inclusive one is always outside the scope of negation
(but see Eckardt 2001 for problems with this generalisation)8.
From a purely conceptual point of view, we find it implausible that the well-known crosslinguistic freedom in the distribution of intensifiers should follow from a conspiracy of
movement operations. In Gast‘s proposal, trying to do so comes at the price of associating each
positional variant of the intensifier with quite ad hoc assumptions about the movement triggers in
the sentences that contain them. What matters most here, however, is the fact that Gast ties
every meaning of the intensifier to a particular structural position, and as illustrated later in this
paper, such an approach accounts only for a fraction of the data.
2.2. One lexical entry, many structural positions (round 1)
The common characteristic of the accounts outlined so far is the fact that they assume a
correlation between the structural position of the intensifier and its interpretation. In other words,
they predict that an intensifier found attached to the nominal DP it interacts with cannot have the
same interpretation as one that is not adnominal. This prediction is immediately falsified by the
following examples, where the continuation and not his secretary ensures that the interpretation
of the intensifier remains stable across the board.
(8)
a. The director himself has appeared, and not his secretary.
b. The director has himself appeared, and not his secretary.
c. The director has appeared himself, and not his secretary.
Note that our informants have explicitly stated that the above examples do not differ in the
slightest either with regard to grammaticality or interpretation. Their interpretation is consistent
6
Note that these are not the only movement operations that Gast assumes. In order to derive the sentence final
inclusive intensifier in Swedish, he further assumes that the whole verb phrase moves across the intensifier to some
position above TP. This type of movement is widely known as “heavy shift”. The Swedish example below shows
this process (taken from Gast (2006: p. 91))
a)
...för att jag3 [t2 har levt i Oslo]1 [t3 själv]2 t1
. . because I have lived in Oslo myself
7
The same point is valid for the adnominal intensifier found next to the subject of the sentence, or for any
instance of the intensifier which is not sentence final (remember that intensifiers are always stressed).
8
Due to lack of space we do not offer a full review of Gast’s analysis.
UCLWPL 2011.01
86
with the widely adopted view that the intensifier evokes alternatives to the referent of the
intensified DP, whilst at the same time excluding them (what König 1991, Siemund 2000 and
Gast 2006, among others, call the adnominal meaning). In the case of (8a-c) the contrasting
alternative is the director’s secretary. What becomes apparent from (8a-c) is that the context, and
not the structural position of the intensifier, is the decisive factor for its interpretation. Note that
if it was the case that this intensifier meaning was intrinsically tied to the adnominal structural
position, the sentences (8b-c) should at least be rendered infelicitous due to the strong effect of
the surrounding context. Indeed, the linguistic context supplied in these examples should be
incompatible with the meaning of the intensifier. The examples below apply the same test to the
remaining two readings of the intensifier that we have encountered up to now (paraphrased as
also and alone/by x-self).
(9)
Context: Speaker A: Being poor is tough.
a. Speaker B: Yes, I know, I myself have been poor and I remember it wasn‘t easy.
b. Speaker B: Yes, I know, I have myself been poor and I remember it wasn‘t easy.
c. Speaker B: Yes, I know, I have been poor myself and I remember it wasn‘t easy.
(10)
Context: Speaker A: John found the way to the station with his brother.
a. Speaker B: ? No, no John himself has found the way to the station.
b. Speaker B: ? No, no John has himself found the way to the station.
c. Speaker B: No, no John has found the way to the station himself.
The examples in (9a-c) illustrate the same point as before but with the also reading of the
intensifier. This reading exhibits the same distributional flexibility found with the in person one.
The relative infelicity of (10a-b), however, is unexpected, given what we concluded with regard
to the other two readings of the intensifier. As evident from (10a-c), the alone reading of the
intensifier seems to prefer a post-verbal position. It cannot be the case that this position is tied to
this reading because we have already seen examples where a post-verbal intensifier does not
have the alone reading. Another idiosyncrasy of this reading is that it imposes restrictions on the
event-type of the predication it is found in. As König (1991) points out, this reading is present
only in sentences describing non-repeatable activities. Siemund (2000) elaborates on this idea
and suggests that the alone/by x-self reading is only compatible with verbs denoting
accomplishments or achievements (the examples in (10) all have interpretations compatible with
this claim). This is illustrated in the example provided by Gast (2006) in (11), which becomes
infelicitous if we attempt to interpret myself in this way (alone). This is due to the stative nature
of the verb to be.
(11) # I am a gardener myself.
In section 3 we argue that both the incompatibility of this reading with stative verbs and its
restriction to post-verbal positions is a consequence of the nature of the evoked alternatives. For
the time being, note that the wider context is a crucial factor in defining the interpretation of the
intensifier. The example below, which repeats (10c) but with a different context, is intended to
show that intra-sentential factors (i.e. type of predicate) do not have anything to do with the
choice of the reading. Hence, this time the preferred reading is the also one.
(12) Speaker A: Bill managed to find the way to the station the other day, surprisingly!
UCLWPL 2011.01
87
Speaker B: Well, he is not the only one! John found the way to the station himself!
The overall discussion up to this point leads us to the conclusion that approaches which correlate
a particular interpretation of the intensifier with a syntactic position are far too restrictive.
Evaluating these accounts from a purely interpretational perspective illustrates the same
point; assuming two or three fixed readings for the intensifier is again far too restrictive. There is
an abundance of interpretations of the intensifier; Cohen (1999) provides us with the example in
(13), in which the intensifier has a reading closely paraphrasable to even (henceforth called the
even reading); a reading akin to the focus particle only is shown in (14) (henceforth called the
only reading), even though it is not perfectly felicitous (see section 3 for an explanation); finally,
a reading similar to the phrase as far as x is concerned (or as for x) (henceforth called the as for
reading) is illustrated in (15)9.
(13) Clinton himself will vote for the Republicans. (paraphrase: Even Clinton will ...)
(14) Speaker A: Mary gave Jane her syntax course-book and all the notes before leaving the
university. At least that what she told me!
Speaker B: ?Well, you must have misheard! Mary gave the syntax course-book itself, and
threw away the notes. (paraphrase: ... Mary gave John only the syntax ...)
(15) Speaker A: How was your first flight to the US? (Speaker B is afraid of flying)
Speaker B: Well, the flight itself was ok, but seriously... you don‘t want to know about the
landing! (paraphrase: Well, as for the flight, it was ok, ...)
Up to this point we have pin-pointed six readings of the intensifier, and there may well be others.
Our intention though is not simply to spell out a full list of the interpretations of the intensifier.
Instead, we attempt (in section 3) to provide a uniform treatment of intensifiers that assumes a
single lexical entry, while being flexible enough to accommodate and predict all the possible
readings of the intensifier.
2.3. One lexical entry, many structural positions (round 2)
Contrary to the empirical character of the arguments presented above, we now attempt to argue
for a common lexical entry for all the uses of the intensifier based on arguments of a more
conceptual basis, which in turn are based on empirical observations found in the literature. The
first argument is based on the (obvious) observation that the various readings of the intensifier
(i.e. himself) do not affect its morphological make-up. For researchers who assume a different
lexical entry for each of these readings, this amounts to the claim that a single word can have at
least six different lexical entries (these are the readings we detected up to this point). On top of
this, the same word serves the purpose of being a reflexive in English, thus adding one more
entry, leaving us with at least seven entries. Clearly, this is theoretically unattractive. Moreover,
to our knowledge, there is no other word in the language with so many lexical entries. Note that
9
The reading is most easily accessible with a B-accent, characteristic of contrastive topics (Jackendoff 1972),
maximally realised as L+H* and followed by a default low tone and a high boundary tone (L H%) on the intensifier.
Contrastive topic is indicated with double underlining throughout the rest of the paper.
UCLWPL 2011.01
88
the situation in English is not unique. Cross-linguistically, numerous languages10 encode all the
different readings of the intensifier with one word (i.e. Dutch: zelf, German: selbst, Greek: o
idios).
Moravcsik (1972), Siemund (2000), and Gast (2006), among others, argue that the
different uses of the intensifier impose dissimilar selectional restrictions on the nominal
constituent it interacts with. No consensus has been reached with regard to this issue. For
example, Siemund (2000) argues that the in person reading does not impose any restrictions on
its head-DP, thus it can denote anything (i.e. [±human])11, whereas the alone reading cannot
interact with DPs denoting anything other than humans and higher animals. The same author
suggests that the also reading can only interact with humans. On the other hand, Gast (2006)
counter-argues the above claims on the basis of examples such as (16) and (17).
(16) Any business that stores sufficiently large amounts of hazardous waste can be a storage
facility even if it is merely storing the waste that it produced itself on the site where it was
produced.
(17) We are works of art, belonging to a world that is itself an aesthetic phenomenon.
Examples (16) and (17) illustrate that no selectional restrictions on the associate DP in terms of
an animacy hierarchy are imposed by the alone and also readings, respectively. The example
below illustrates that the even reading does not impose any animacy-related selectional
restrictions on its associate DP either.
(18) Context: Two geologists are discussing the fact that Japan’s earthquake was so strong that
it was supposedly felt all over the world. It is shared knowledge between them that France
is the most distant country from Japan.
Japan‘s earthquake was massive! The whole world felt the shake. France itself felt the hit.
Examples (14) – (18) force us to conclude that the different readings of the intensifier do not
impose any restrictions with regard to the animacy status of their head-DP12. This common
property of all readings of the intensifier provides more support for an account assuming a single
lexical entry.
10
We have not been able to find a study that checks this parameter. We suspect from König and Siemund’s
cross-linguistic morphological comparison between intensifiers and reflexives (WALS, 27.3.2011) that most
languages encode the various readings of the intensifier in one word.
11
There is consensus for the in person reading’s absence of animacy-related restrictions. Irrespectively of this
point, we take it for granted that everyone agrees that all the readings can interact with human referents, and
therefore do not provide any examples of such uses.
12
Of course it is not expected that the intensifier will exhibit the same behavior with regard to animacy
restrictions on its associated DP universally. Gast (2009) discusses this issue and illustrates that in Japanese there is
one intensifier that can interact only with animate DPs, namely jishin, and one intensifier that can interact only with
inanimate DPs, namely jitai. As he explains, the reason for this discrepancy is associated with the historical origin of
each word. For example, jishin is lexically derived from the word body, hence its use is restricted to animate DPs
(things which have bodies). What we do expect however, is for one and the same morphological word not to have
different selectional restrictions across different readings.
UCLWPL 2011.01
89
Aside from the animacy restrictions just discussed, there has been no consensus in relation
to whether the intensifier can interact with indefinite DPs, and if so, on which readings.
Moravcsik (1972) provides the minimal pair example below aiming to illustrate that an in person
reading is not able to interact with indefinites, while an alone reading is.
(19) a. An engineer should know this himself.
b. ?An engineer himself should know this.
We believe that examples such as the above have misled various researchers towards accounting
for the wrong facts. Eckardt (2001), for example, takes this picture for granted and assumes a
different semantics for each variant (head-adjacent vs head-distant), particularly tailored to
express that the head-distant (or adverbial) occurrences can interact with indefinites whereas the
head-adjacent ones cannot. Following Siemund (2000), we would like to argue that, despite
appearances, all the readings are equally incapable of interacting with true indefinites. The word
true is meant to exclude indefinite DPs which have a generic or specific interpretation. As
pointed out by Siemund, the apparent ability of the alone reading to interact with indefinites is
owed to the fact that these DPs are invariably interpreted as generic or specific13. Indeed, (19a)
has a generic reading. (20) illustrates that despite its indefinite nature, a general must be
interpreted as specific in the sense that a particular general is under discussion (Siemund, 2000).
(20) A general has commanded the army himself.
This conclusion is in agreement with the fact that regardless of the reading of the intensifier, it
cannot interact with the indefinite pronoun someone as long as it is not read specifically. This is
exemplified below for the alone reading ((21) is provided by Siemund (2000)).
(21) ? Someone wrote me a letter himself.
Note that the example in (19b) can be rendered felicitous when a suitable context is provided that
forces the indefinite‘s interpretation to become generic, as shown in (22).
(22) Context: Discussion about what responsibilities engineers and their assistants are supposed
to have
Speaker A: I think that an engineer‘s assistant is responsible for knowing what new research
is being developed.
Speaker B: No, I disagree, I think an engineer himself should know this.
The examples below illustrate the same point for the also (23), even (23), and as for (24)
readings. The only way for (23) - (24) to be felicitous is if the DP in question is interpreted as a
generic or specific indefinite. Note that in (23) the intensifier can be an instance of either the also
or the even reading.
(23) a. Context: The whole country was surprised by the recent political scandals.
13
We remain agnostic as to why this reading is less contextually dependent compared to the rest of the
readings in terms of its ability to achieve a generic/specific interpretation on the indefinite.
UCLWPL 2011.01
90
A politician was surprised himself. (generic/specific reading)
b. #A politician was surprised himself. (true indefinite reading)
(24) a. Context: Discussing about which animals can be pets.
Speaker A: Even though the dog is a species that comes from the wolf, it can easily be a
pet.
Speaker B: Well, I don‘t know whether this is the case for every dog breed, but a wolf
itself can certainly not be someone‘s pet. (generic reading).
b. # A wolf itself can certainly live less than a hundred years. (true indefinite reading)
Again, we reach the same conclusion as with the animacy parameter. There is no variation
between the intensifier readings as far as the definiteness nature of the head-DP is concerned.
None of them can interact with a true indefinite DP; a commonality that on the one hand again
points towards one lexical entry and on the other indicates a fundamental property of the
intensifier (the centrality effect). This is discussed in the next section, the core proposal of the
paper14.
3. Deriving the various readings of the intensifier (the proposal)
We propose that an intensifier and its associate DP always fulfil the information-structural role
of a topic or a focus. Furthermore, it may or may not be interpreted contrastively, just like topics
or foci that lack an intensifier. The presence of the intensifier gives rise to an additional
interpretive effect though: the characterisation of the value denoted by the associate DP as
central. In section 3.1. we discuss, and adopt, Eckardt‘s (2001) insight that the intensifier denotes
a truth-conditionally meaningless function ID (x). The exact predictions of our proposal depend
on one‘s view of information structure and particularly the interpretations of topics, foci and
contrast. Hence in section 3.2. we outline Neeleman and Vermeulen‘s (2010) approach towards
these notions, which is (slightly modified but essentially) adopted in this paper. In sections 3.3. 3.10. we discuss how the various readings of the intensifier-associate DP constructions fit with
the information structural typology of Neeleman and Vermeulen (2010) and reach the conclusion
that there is only one lexical entry for the intensifier. The intensifier‘s various readings boil down
to the nature of its information-structural role (mainly), which in turn is determined by the
context. Section 3.11. attempts to provide an explanation of why the intensifier is incompatible
with true indefinites.
3.1. The meaning of the intensifier
14
Note that, contrary to the rest of the readings of the intensifier we discovered up to this point, the only
reading is not felicitous with any type of indefinite DP . This is shown below.
a)
Speaker A: I want to buy a pencil and a pen.
Speaker B: #I think you should buy a pen itself. (specific/generic reading)
This, however, does not create any problems for our attempt to unify all the readings of the intensifier under one
lexical entry because, as will be discussed in detail later in the paper, what we call the only reading is not an
intensifier reading to begin with.
UCLWPL 2011.01
91
In an attempt to explain why the intensifier consistently evokes a set of alternatives to the
associate DP that it is intuitively linked to, Eckardt (2001) proposes that it is the intensifier‘s
interaction with focus, and not a property of the intensifier itself, that is responsible for it. In
particular, Eckardt, who follows Moravcsik (1972), suggests that the core meaning contribution
of the intensifier is the identity function ID on the domain of objects De.
(25) ID: De De
ID (α) = α for all α ∈ De
According to this analysis, the intensifier is merely lexically specified with ID, which takes as its
input value a nominal constituent x, the associate DP, and maps it onto the same output value.
(26) exemplifies this operation for the DP John himself.
(26) 〚[John] himself〛 = ID (〚John〛) = 〚John〛
Adopting the assumption that the intensifier denotes ID is equivalent to saying that its core
meaning contribution to the sentence amounts to nil. We agree with this view mainly for two
reasons. One, it makes perfect sense from an interpretive perspective; the DP John himself does
not have a different denotation from John. Two, it predicts obligatory stress on the intensifier.
Eckardt proposes that the obligatory stress indicates that the constituent is in focus, and like
every other focused constituent, it evokes alternatives, contributing in this way to the meaning of
the sentence; hence, the invariable presence of alternatives in intensifier constructions15.
However, this picture is overly simplistic. Eckardt takes it for granted that the stress on the
intensifier is usually associated with emphatic focus (hence the surprise inference), whilst we
have seen examples in which an intensifier and its associate fulfil another information-structural
role, namely that of contrastive topic (see example (15)). Even though she notes that such a
reading (with a hat contour accent) is indeed possible, she only states that this use of the
intensifier, along with some other uses (i.e. in contexts of question-answer focus and of use of
the focus particle only), does not express surprise. What would be more interesting though is to
find out what this reading does express and how it is achieved, as well as to figure out a way of
unifying all the no-surprise cases (as Eckardt calls them) with the surprise ones. The stand we
take in this paper suggests that the notion of surprise is simply redundant in accounting for the
full range of uses of the intensifier. Instead, we need the notions of contrast, topic and focus, as
described by Neeleman and Vermeulen (2010), in order to capture the underlying mechanisms
that govern the possible readings of the intensifier.
We propose that the intensifier and its associate DP can potentially take the form of any
notion, or a possible combination of the notions, of the component of information structure,16
15
It has to be noted that Eckardt makes the crucial assumption that the intensifier is in focus and the associate
DP in its propositional background and it is this kind of relationship that eventually gives us alternatives to the
associated DP and not alternatives to the intensifier itself. The technical details of this process are discussed later in
this paper.
16
As Neeleman and Vermeulen (2010) point out, these notions can be of a discourse nature targeted by
mapping rules operating between different components of grammar (i.e. information structure and syntax). As far as
we can see, this does not create any issues for our approach here.
UCLWPL 2011.01
92
which comprises [contrast], [topic] and [focus]. These notions are organised systematically as
shown in the table below, (first presented by Neeleman and Van de Koot 2008)).
(27)
Contrast
Topic
Focus
Aboutness topic
[Topic]
New information focus
[Focus]
Contrastive Topic
[Topic, Contrast]
Contrastive Focus
[Focus, Contrast]
The table expresses that topic and focus are basic notions of information structure that can be
enriched to yield a contrastive interpretation. This results in the following four way typology of
information structural categories; focus, topic, contrastive focus, contrastive topic. The
independent existence of these categories and their linguistic relevance have been extensively
argued in various works (Reinhart 1981; Rizzi, 1997; Kiss 1998; Vallduvi and Vilkuna, 1998;
Molnar, 2002; Neeleman and van De Koot, 2008; Neeleman and Vermeulen (2010); among
others). As will be illustrated, the interpretive characteristics of these categories constitute the
basic interpretive aspects of the various readings of the intensifier.
3.2. The interpretation of [contrast], [topic] and [focus]
The Selfish Gene in (28B) is commonly assumed to be in focus because it corresponds to the whexpression found in (28A).
(28) Speaker A: What did John read?
Speaker B: He read [The Selfish Gene]F.
As pointed out by Selkirk (1984, 1996) and others, the focused constituent receives the main
stress of the sentence. Following Rooth (1985, 1992), Neeleman and Vermeulen (2010) suggest
that the focused constituent evokes a set of alternative propositions that differ only in the focused
position and share all the rest of the material, the focus value of the sentence. The ordinary value
of the sentence is the proposition expressed by the sentence. Below are the ordinary and focus
values of (28B).
(29) Ordinary value: [John read The Selfish Gene]
Focus value: {[John read The Selfish Gene], [John read The Blind Watchmaker], [John
read The Ancestor‘s Tale], [John read The Extended Phenotype],...}
The information in (29) can be represented in the somewhat different notational variant of (30)
provided by Neeleman and Vermeulen (2010), which we adopt in this paper for reasons of
simplicity.
(30) <λx [John read x], The Selfish Gene, {The Selfish Gene, The Blind Watchmaker, The
Ancestor‘s Tale, The Extended Phenotype,...}>
UCLWPL 2011.01
93
When (28B) is compared to (31B) below, there is an interpretive difference17. Whereas in the
latter example the focused constituent stands in opposition to an alternative explicitly mentioned
in the discourse, in the former there is no explicit alternative and no sense of contrast (Neeleman
and Vermeulen, 2010).
(31) Speaker A: John read The Extended Phenotype.
Speaker B: (No, you‘re wrong) THE SELFISH GENE he read.
(31B) is an instance of a proposition containing a constituent which is interpretatively a
combination of the notions of focus and contrast, a contrastive focus. Neeleman and Vermeulen
(2010) propose that contrast corresponds to a quantifier which gives information about the
relation between two sets, similarly to every other quantifier (i.e. every, some). On this view,
contrast in (31B) expresses to what extent the set α of contextually relevant books is contained in
the set β of things that John read. Two assertions are made: a) one member of α is also a member
of β, and b) there is at least one other member of α that is not contained in β (The Extended
Phenotype). The presence of alternatives and the positive statement in (a) are a result of the
semantics of focus, whereas the negative statement in (b) is a result of the semantics of contrast.
Therefore, contrastive focus and regular focus differ in that only the former encodes a negative
statement. The interpretation of (31B) is shown below in (32).
(32) a. <λx [John read x], The Selfish Gene, {The Selfish Gene, The Blind Watchmaker,
The Ancestor‘s Tale, The Extended Phenotype,...}>
b. x [x ∈ {The Selfish Gene, The Blind Watchmaker, The Ancestor‘s Tale, The
Extended Phenotype,...} &  [John read x]]
Contrary to what is the case with the notion of focus, researchers have not reached a consensus
with respect to the content and linguistic relevance of the notion of topic (compare Chafe 1976,
Reinhart 1981, Vallduvi 1992, Lambrecht 1994). We follow Neeleman and Van de Koot (2008)
and Neeleman and Vermeulen (2010), who in turn follow Reinhart (1981), in characterizing
topics in terms of ―aboutness‖. Note that Neeleman and Vermeulen (2010) draw a clear
distinction between ‗discourse topics‘ and ‗sentence topics‘. Discourse topics refer to entities
that a unit of discourse is about, whereas sentence topics refer to the syntactic constituents used
to introduce a referent that the sentence is about. Since the notion of discourse topic is not
directly relevant to this paper, henceforth we refer to the notion of topic as to mean sentence
topic. The subject in (33B) is an instance of topic.
(33) Speaker A: Tell me about one of your friends.
Speaker B: Well, [Maxine]T was invited to a party by Claire on her first trip to New
York.
Similarly to foci, topics are associated with a set of alternatives. However, contrary to the
propositional nature of the focus alternatives, the lambda operator generates utterances. The
representation of topic differs from that of focus in that the function contains an assertion
operator, which means that its application derives utterances rather than propositions. The
17
Instances of contrastive focus are represented with SMALL CAPS throughout the rest of the paper. According
to Jackendoff (1972), contrastive focus requires an A-accent in English, a plain high tone (H*) often followed by a
default low tone. Regular focus on objects is usually marked with nuclear stress.
UCLWPL 2011.01
94
representation of the ordinary value and topic value (in parallel to the focus value) of (33B) is
shown below as (34). Note that the representation below is in accordance to the intuition that the
speaker performs the following speech acts when uttering (33B): a) Consider Maxine (out of a
set of possible topics); b) I assert that Maxine was invited by Claire to a party in New York.
(34) <λx ASSERT [x was invited by Claire to a party in New York], Maxine, {Maxine,
Susan, Bill,...}>
Similarly to the notion of focus, the notion of topic can also be interpreted contrastively. In (35),
Maxine stands in opposition to an alternative explicitly mentioned in the discourse, Bill.
(35) Speaker A: Tell me about Bill. Was he invited to a party when he went to New York?
Speaker B: Well, I don‘t know about Bill, but Maxine was invited to a party on her first
trip to New York by Claire.
Since the alternatives evoked by topics (and contrastive topics) are utterances, and not
propositions as is the case for focus, the interpretational effect associated with contrast is that the
speaker is unwilling to make (at least) one alternative assertion. As Vermeulen (2010) points out,
since contrastive topic is an utterance level notion, the reason for not committing to an
alternative utterance must be pragmatic (i.e. the speaker does not want to be held responsible for
the information conveyed by the relevant alternative). In a nutshell, Neeleman and Vermeulen
(2010) suggest that contrastive foci deny at least an alternative proposition, whereas contrastive
topics indicate that the speaker is unwilling (for a pragmatic reason) to make an alternative
utterance. (36) constitutes the interpretation of (35B).
(36) a. <λx ASSERT [x was invited by Claire to a party in New York], Maxine, {Maxine,
Susan, Bill,...}>
b. y [y ∈ {Maxine, Bill, Susan,...} & (y λx ASSERT [x was invited by Claire to a
party in New York])].
Equipped with the interpretation of the notions of contrast, topic and focus, and their
possible combinations, we are now able to derive all the readings of the intensifier. As
previously suggested, we hypothesize that the intensifier and its associate can carry these
information-structural interpretations (i.e. focus, topic), or a combination of them (i.e. contrastive
focus, contrastive topic). We will argue that each reading of the intensifier can be explicated in
terms of the interpretive characteristics of one of these notions. The analysis that follows
concentrates on the English intensifier, and as we will see the approach advocated here not only
captures the readings already discussed, but makes predictions regarding the possible readings of
an intensifier. Importantly, these predictions are borne out by the discovery of readings that, to
our knowledge, have gone unnoticed (i.e. the reading marked as topic).
3.3. The in person reading
The reading of the intensifier that has undoubtedly attracted most interest in the prior literature is
the in person one. Suppose the intensifier is associated with an argument a and that P is the
predicate resulting from performing lambda abstraction on a. Then the meaning contribution of
the in person reading can be summarized as follows; a) it evokes a set of alternative referents
that includes a, b) it structures this set into a central element a and peripheral elements {b, …}
UCLWPL 2011.01
95
(König and Siemund 1999, Siemund 2000, Eckardt 2001, Hole 2002, Gast 2006, among others),
and c) (P(a) is true, while application of λx.P(x) yields a false proposition for (at least) one of the
evoked alternatives. These characteristics are now discussed in more detail. (37) contains an
instance of the intensifier with the in person reading. The intensified value is the direct object.
(37) Yesterday, I saw [DP John himself].
An important property of the intensifier construction in (37) is the evoking of a set of alternative
sentences, which are formed by replacing John himself with alternative DPs. Even though it
seems tempting to assume that the DP John himself is just like every other constituent that is
focused, such an analysis runs into problems when considering (28) repeated below as (38).
(38) Speaker A: What did John read?
Speaker B: He read [The Selfish Gene]F.
There are two main differences between the sets of alternatives of (37) and (38B), as first pointed
out by Gast (2006). The first one has to do with the saliency of these alternatives. The intensifier
in (37) invariably makes reference to contextually given alternatives, which are evoked if not
present (i.e. Yesterday, I saw [DP the brother of John]). This is not the case for (38B), which does
not evoke any specific alternatives (i.e. He read [The Extended Phenotype] F). Thus, unlike what
is the case for (37), the person uttering (38B) does not need to have any particular alternatives in
mind. The second difference has to do with the nature of the evoked alternatives, namely the fact
that in the case of (37), the alternatives must have something to do with the associate DP.
Moreover, the associate‘s referent must be central with respect to the referents denoted by the
alternative values, the entourage. This requirement is exemplified in (39). Assuming that there is
not some sort of relation holding between Mary and John, the option of identifying the referent
of the alternative value (Mary) as peripheral to the referent of the intensified DP is ruled out;
hence Speaker B‘s reply in (39b) is infelicitous. On the other hand, the use of the intensifier in
(39a) is felicitous because the referent of the alternative value (the brother of John) is identified
through the intensified referent, John, by the use of the relational noun brother of18, hence the
understanding of John as being central.
(39) a. Speaker A: Bill told me that you saw the brother of John yesterday.
Speaker B: No, no, yesterday I saw John himself.
b. Speaker A: Bill told me that you saw Mary yesterday.
Speaker B: #No, no, yesterday I saw John himself.
As Gast (2006) points out, the centrality effect imposed by the intensifier on its associate DP is
not found in focus constructions. Focused constituents allow reference to alternatives that are
only restricted in terms of their semantic type. The contrast between a constituent which is
intensified and one which is focused is also visible in the example provided by Gast (2006) in
(40), once we assume that there is no historical or contextual relationship between the two
islands, in which Hawaii is central and Tahiti peripheral (i.e. Tahiti is financially dependent on
Hawaii).
18
Note that the use of relational nouns, such as bother of, sister of etc is not a prerequisite for the felicitous use
of the intensifier. See Siemund (2000) for a list of contexts that fulfil the requirements of the felicitous use of the in
person reading.
UCLWPL 2011.01
96
(40) a. #I have never been to Hawaii itself, but I‘ve been to Tahiti.
b. I have never been to [Hawaii]F, but I‘ve been to [Tahiti]F.
What can be concluded from the examples above is that the intensifier can only associate
with values that have been previously rendered as central in discourse, in some way. In (39a) for
instance, the alternative value brother of John is a possessive construction consisting of two
entities, namely the brother and John. As Nikolaeva and Spencer (2010) point out, the
relationship between the possessor and the possessee is largely asymmetric. Following various
authors, they further point out that the possessors function as pragmatic anchors (J. A. Hawkins
1978; R. Hawkins 1981; Fraurud 1990; Koptjevskaja-Tamm 2000, 2004) or reference points for
identifying the possessee (Langacker 1993, 1995). For (40a) we pointed out that the intensifier
can be used felicitously only if a similar asymmetric relationship (i.e. Tahiti is dependent in
some way on Hawaii) holds, a priori, between the intensified value and its alternative.
Apart from structuring a set of salient alternatives into a centre and periphery, the in person
reading exhibits one last significant function. It opposes the referent of the intensified DP to the
referents of the evoked alternative values, and, importantly, excludes at least19 one of the
corresponding alternative propositions. (41) means that the president, and not someone else
related to him (i.e. the president‘s secretary, the vice-president), will perform the action
described by the predicate.
(41) The president himself will announce the decision of the cabinet (and not his secretary).
Again, the exclusion of alternatives is not necessarily found in focus constructions. Speaker B‘s
reply in (28), repeated below as (42), is not interpreted as excluding an alternative (i.e. John read
[The Extended Phenotype]F).
(42) Speaker A: What did John read?
Speaker B: He read the [The Selfish Gene]F.
Taking into account the interpretation associated with the information structural categories
predicted by the typological table in (27) and the various aspects of the meaning contribution of
the in person reading, we may hypothesize the following:
(43) Hypothesis 1 (first draft): The in person reading of the intensifier is invariably an
instance of contrastive focus.
19
We stress the point that the in person reading is not understood exhaustively (in the sense of Kiss 1998, i.e.
negates all alternatives), but only contrastively. After all, if it were the case that this reading is characterized with
exhaustivity the example below should be, but is not, infelicitous. As can be seen, someone peripheral to Mary, her
brother, is not excluded from the action denoted by the predicate. The fronting of the intensifier-DP construction
ensures that it is interpreted contrastively (see Neeleman and Van de Koot 2008). This is in accord to our hypothesis
below that the in person reading and its associate DP are interpreted as contrastive focus.
a)
Context: John intends to marry Mary. However, it is tradition that he has to meet her whole family in order
to be approved by them.
John’s mother: John met Mary yesterday!
John’s father: Yes, Mary HERSELF he met, her BROTHER he also met, but he didn’t meet the rest of her
family.
UCLWPL 2011.01
97
Given that the intensifier i) denotes the function ID (x) (that maps its input to the same output),
ii) is satisfied by a nominal constituent, namely the associate DP that it is intuitively understood
to interact with, iii) is (contrastively) focused and iv) that the effect of such focusing is the
inducing of alternatives of the same semantic type as the asserted value (Rooth, 1985, 1992), the
alternative values that are induced should also be functions from De to De. In the case of the in
person reading, we assume that ID (x) contrasts with the function DEP (x), which is paraphrased
as an entity dependent on (x) (see Eckardt 2001, Hole 2002, Gast 2006 for similar approaches).
Let us call this function the dependent function. This function, by definition, restricts the
alternative values to x (x = the value denoted by the nominal constituent interacting with the
intensifier) to only those that are peripheral to x, because it encompasses the asymmetrical
relationship holding between the entity x and some other entity in the alternative value. Contrary
to ID (x) though, DEP (x) does not map its input onto the same output, when applied to x (in the
same fashion as ID (x)). Assuming that one of the evoked alternatives to the DP Hawaii itself is
Tahiti, DEP takes Hawaii as its input and maps it onto an entity dependent on Hawaii, which in
this case is Tahiti. This process is exemplified below.
(44) 〚Tahiti〛 = DEP (〚Hawaii〛) = 〚an entity dependent on Hawaii〛 = 〚Tahiti〛
Note that our approach is significantly different from authors (Eckardt 2001, Hole 2002 among
others) who assume that the centrality effect is a result of the fact that the alternative values are
structured around the intensified value. To make things more transparent, these authors assume
that the centrality effect simply results from the presence of the intensified value in the
alternative values. It is true that the alternative values invariably consist of the intensified entity
(and some other entity). This can be seen in (44), in which the alternative value of Hawaii,
Tahiti, is indeed defined in terms of Hawaii, through the application of DEP. However, despite
the attractiveness of this approach, it falls short in expressing the fact that the intensified value
needs a specific type of alternatives (the entities found in the alternative values must be a priori
structured in a centre-periphery fashion). The mere presence of the intensified entity within all of
its alternative values is not enough. In fact, this assumption makes the wrong predictions. As we
will see in section 3.6 an explanation of centrality along these lines over-generates, in the sense
that it predicts impossible alternative values for the intensified value, and hence readings of the
intensifier that do not exist. On the other hand, our approach suggests that the alternative values
are structured around a central element x beforehand (intensification) with various means such as
the use of the possessive construction or historical events that make a country dependent on
some other one (see example (40a) and surrounding discussion). This asymmetric relation
between the entity that eventually becomes the intensified value and another entity, both forming
part of the alternative value, is precisely what DEP expresses. Of course, for this asymmetric
relation (or for any relation) to exist, both entities need to be present in what we call the
alternative value. Therefore, the consistent presence of the intensified value in the alternative
values is merely an epiphenomenon. Notice that the assumption that ID (x) contrasts only with
DEP (x) predicts that the intensifier will only associate with DPs whose denotations are central,
hence the in person reading‘s observed centrality effect is explained. For the sake of
completeness we would like to point out that, on a par of the possessive construction, there are
other ways of expressing an asymmetric relation between two entities that constitute part of one
syntactic constituent, and hence are felicitous alternatives to the in person reading and its
UCLWPL 2011.01
98
associate DP. The example below illustrates this. The subject is a complex DP consisting of the
head DP and its modifier. The modifier, near the table, functions as a reference point for the
identification of the head DP, the chair. This is reminiscent (see above) of the role played by
(and hence the asymmetric relation between) the entities in a possessive construction. However,
contrary to the asymmetric relation expressed by possession, the relation holding between the
entities in the subject constituent of the example below is a relation of distance, as expressed by
the preposition near. The fact that we are able to define, with no particular difficulty, the nature
of the relation expressed by near will become particularly important later on, when we will
discuss the exact nature of the entities comprising the alternative referents of the intensified DP.
(45) Speaker A: The chair near the table looks nice.
Speaker B: Well, I think that the table ITSELF looks nice. The chair is awful.
We have not yet explained though the requirement of this reading to interact with salient
alternatives and the effect of excluding them (as will become more explicit in the representations
given in (47) and (48), this reading does not interact with, and exclude, the salient alternative
values themselves but alternative propositions involving these alternative values). This is where
the notion of contrast comes into play. As pointed out above, contrast has two main functions.
One, it refers to alternatives explicitly mentioned in discourse or evokes them if not present; in
other words it requires salient alternatives. Two, when combined with the notion of focus, it
excludes (at least one of) these alternatives, which have the form of propositions. Based on these
parallel interpretational effects, it seems reasonable to analyse the DP that interacts with an
intensifier with the in person reading along the lines of a contrastively focused constituent.
A note is in order here with respect to (43). (43) states that it is the intensifier itself, and not
the DP it interacts with, that is contrastively focused marked. This means that, contrary to a
regularly contrastively focused marked constituent, the intensified nominal constituent is not
marked as such. The interpretation of this constituent as a contrastive focus occurs indirectly
through its interaction with the intensifier (and as we have seen it is this indirect route that also
gives rise to the centrality effect). The representation of (37), repeated below as (46) without
yesterday for the sake of simplicity, is shown in (47)20.
(46) I saw John HIMSELF.
(47) a. <λx [I saw x], John, {John, DEP_1_(John), DEP_2_(John), ..., DEP_n_(John)}>
b. x [x ∈ {John, DEP_1_ (John), DEP_2_(John), ..., DEP_n_(John)} &  [I saw x]]
In order to see more clearly the effect of context with regard to restricting the set of alternatives,
the representation of speaker B‘s utterance in (39a) is provided below in (48). This time the set
of referents is closed and consists only of two entities as specified by the context. As the
semantics illustrate below, the brother of John is represented as DEP (John).
20
We will follow this type of representations throughout the rest of the paper. However, we are aware that this
representation disregards issues having to do with the position of the intensifier. This representation and every one
that follows essentially take it for granted that the intensifier always takes scope over the entire proposition/utterance
(as there are never variables that are closed off through existential closure), something which cannot always be right.
Further research is needed in order to determine whether the intensifier is able to take different scope from the one
determined by its position in surface syntax, and if this is possible, whether there are any restrictions in this process.
UCLWPL 2011.01
99
(48) a. <λx [I saw x], John, {John, DEP (John)}>
b. x [x ∈ {John, DEP (John)} &  [I saw x]]
(48) expresses to what extent the set of contextually relevant entities (i.e. John, the brother of
John, etc) is contained in the things that I saw. It is asserted that one member of this set of
entities is also a member of the set of things that I saw. It is also asserted that there is at least one
other member of this set that is not contained in the set of things I saw (i.e. the brother of John).
Similarly to a constituent marked with contrastive focus, the presence of alternatives and the
positive statement are a result of the semantics of focus, whereas the negative statement is a
result of the semantics of contrast. Contrary to just any other contrastively focused constituent
however, the presence of the in person reading in (39a) and (46) constrains the members of the
set of alternatives in such a way that only certain variants of the proposition I saw John can be
included. The position held by John in these variants can only be filled by constituents denoting
referents peripheral to John (i.e. the brother of John), and nothing else.
All things considered, the meaning contribution of an intensifier carrying the in person
reading is crucially a result of the interpretational effects of contrastive focus interacting with ID
(x). This interaction evokes a particular type of alternatives, which are assumed to be the
dependent function(s). We may therefore replace (43) with the more articulated hypothesis in
(49).
(49) Hypothesis 1 (second and final draft): The in person reading of the intensifier is invariably
an instance of contrastive focus and always makes reference to dependent function(s).
It is now predicted that the felicity conditions of the in person reading should show parallel
behaviour to the conditions of a contrastively focused constituent. A comparison between (50)
and (51) illustrates that this prediction is borne out.
(50) Speaker A: What did John read?
Speaker B: #THE SELFISH GENE he read.
(51) Speaker A: Who did you see yesterday?
Speaker B: #I saw John HIMSELF, yesterday.
As Neeleman and Vermeulen (2010) point out, speaker B‘s answer in (50) is infelicitous in that
context because the contrast implied by The Selfish Gene and some other reading material cannot
easily be accommodated by speaker A. This is because the alternative reading material is not
made accessible to him/her in the discourse. As a result, speaker B‘s answer is likely to trigger a
request for clarification, such as What do you mean? What did he not read? It has to be noted
here that the requirement for availability of salient alternatives is intrinsically connected to the
main function of contrast, which is to negate alternatives. In order to successfully negate
alternatives, the speaker/hearer must have them under consideration. Therefore, the requirement
for salient alternatives in contrastive contexts can be seen as a by-product of the function of
contrast to negate them. The situation in (51) is similar to that of (50). Speaker B‘s reply is
infelicitous because the contrast implied by the DP John himself and someone else related to
John cannot easily be accommodated by speaker A. This is due to the unavailability of salient
alternatives that the contrastively focused intensifier and its associate DP need to make reference
UCLWPL 2011.01
100
to. However, as pointed out above, if these alternatives are not salient, the intensifier evokes
them. Thus, speaker B‘s reply will trigger a request for clarification along the lines of What do
you mean? Who didn’t you see (related to John)?
The analysis proposed for the in person reading also accounts for another reading of the
intensifier, the delegative. Its use is understood as ―the intensified referent‖ has done the action
denoted by the predication instead of having it done by someone else. In the example below for
instance, Mary is understood to have done the action of dying her hair instead of having someone
else (i.e. the hairdresser) performing the action for her.
(52) Speaker A: The hairdresser dyed Mary‘s hair.
Speaker B: No, Mary has dyed her hair HERSELF (and not the hairdresser).
Similarly to the in person reading, the proposition in (52B) can have a continuation of the sort
and not x. This is because it is understood as negating an alternative proposition that differs with
the one asserted only with respect to the position occupied by the intensified referent (and the
absence of the intensifier). Hence, the proposition in (52B) contrasts with a proposition of the
sort in (52A). The question that arises then is how do we get the extra inference in the case of the
delegative reading, namely that the alternative referent contained in the excluded alternative
proposition performs an action on behalf of the intensified referent. We think that this is clearly a
result of pragmatics. The delegative reading comes across only when the intensified referent has
a direct interest in the action described by the predicate. In the case of (52B), Mary has a direct
interest in the action of dying her hair. Note that the delegative reading is always accompanied
in the predication by a possessive DP of the type her hair (where her refers to the intensified
referent). The presence of such DPs is what expresses the direct interest of the intensified
referent in the denoted action. Once we attempt to change the referent of her in (52), as in (53)
below, the delegative inference becomes unavailable. Instead, as expected, the in person reading
arises.
(53) Mary has dyed Jane‘s hair HERSELF, and not the hairdresser.
Another contextual factor that decides among the two readings is the type of the verb that is
present in the predication. According to Eckardt (2001), the use of the delegative reading is
restricted to combination with agentive verbs (or verbs denoting an action). As shown in (54),
the intensifier can only be an instance of the in person reading (again as expected when the
delegative inference is removed). This is because of the stative nature of the verb.
(54) John is a gardener HIMSELF (not his brother).
This restriction is expected once we consider the interpretative contribution of the delegative
reading (see above). It is inconceivable for someone Y to be a gardener for someone X. We can
think of the delegative reading as a way of transferring whatever is denoted by the predication,
from Y (the alternative referent) to X (the intensified referent). The only conceivable things that
can be transferred and simultaneously be denoted by verbs are actions, hence it is restricted to
occur only with agentive verbs.
In the same way that the in person reading requires that the intensified referent is central,
the delegative reading requires Mary in (52) to be central and the hairdresser as the entourage.
UCLWPL 2011.01
101
This kind of asymmetric relation is defined against the denotation of the rest of the predication,
which is about dying Mary‘s hair. Therefore, we suggest the same analysis for the delegative
reading as for the in person one. ID (x) is contrastively focused and evokes the dependent
function (DEP (x)), whose application on the intensified referent will deliver the hairdresser in
the alternative (and excluded) proposition of (52B), in (52A). This process is shown below21.
(55) 〚the hairdresser〛 = DEP (〚Mary〛) = 〚an entity dependent on Mary〛 = 〚the
hairdresser〛
The example in (56) further substantiates that the delegative reading and its associate DP are
interpreted as contrastive focus, as its infelicity in a context of question-answer focus illustrates
(for those who find this example felicitous, it is because there is evocation of alternative values
that include something like the hairdresser. This further shows that this reading must be
interpreted contrastively in order to be felicitous).
(56) Speaker A: Who has dyed Mary‘s hair?
Speaker B: #Mary has dyed her hair HERSELF.
3.4. The alone reading
Example (57) contains an intensifier with the alone reading. Its use gives rise to the inference
that the subject referent performs the action described by the predicate alone or without help
(hence widely called the assistive reading in the literature), hence the felicitous (albeit slightly
redundant) continuation of the sentence without Bill.
(57) John found the way to the station himself (without Bill).
In contrast to what happens with a simple new information focus (see the discussion surrounding
(42), the intensifier in (57) forces us to consider someone that did not participate together with
John in the action of finding the way to the station (i.e. someone who usually gives directions to
John). To put it otherwise, the main contribution of this reading is to state that the intensified
referent has acted without the external input of some other agent. This fact points towards a
constituent whose function is to evoke alternative referents (to its associate DP) and at the same
time to exclude them from the action described by the predicate. In fact, we can be more specific
than that, and say that the sentence containing the alone reading contrasts with, and negates, a
salient alternative proposition stating that the action denoted by the proposition containing the
intensifier is carried out by the intensified referent together with some other entity. This means
that (57) contrasts with, and negates, something like (58 below.
(58) John found the way to the station with Bill.
Following this reasoning, we conclude that the alone reading along with its associate DP is an
instance of contrastive focus.
21
To save space and to avoid repetition, we do not provide a formalised analysis of this reading. The reader should
assume that it is the same as with the in person reading.
UCLWPL 2011.01
102
However, at first sight the centrality status required by this reading on its associate DP
differs from the one required by the in person one. As pointed out above, the in person reading
requires the entity that is subsequently intensified to be related in some asymmetrical way (i.e.
through kinship relations, financial dependence, distance) with another entity in the alternative
value. This, we argued, led the way for the intensified value to be understood as central within
the alternative value. On the other hand, the alone reading does not seem to impose this kind of
requirement on the entities being present in discourse. Even if we consider the most extreme
situation in which John and Bill do not know each other, the example in (57) remains felicitous.
In order to understand why this is the case, we need to take a closer look at what the alternative
values of the intensified value in (57) look like.
Once we agree that (57) contrasts with something like (58), the alternative value of the
intensifier and its associate DP John in (57) must be John with Bill and not simply Bill. If it were
just Bill, then we should be able to have a continuation of the sort and not Bill. However, this
continuation does not give rise to the targeted reading but the in person one. The only felicitous
continuation of propositions containing the alone reading of the intensifier must consist of
something that means, or is semantically equivalent to, without y (i.e. not in the company of y).
Two things can be concluded from the above discussion. One, the alternative value of the
alone reading and its associate DP x is x with y. Crucially, it is not just y. Two, the alternative
propositions evoked with the use of the alone reading are invariably an instance of a concomitant
construction. Concomitant constructions are the linguistic means of expressing a relation
between the entities/participants of the event denoted by a proposition. According to Lehmann
and Shin (2005), the concomitant, or comitative, construction (i.e. with y) can be characterised as
a relator joining x and y, where y is a participant (henceforth called concomitant participant) that
accompanies or associates with x, which is another participant, usually the actor. The same
authors further point out that the general meaning of a comitative construction consists of the
following features (C = concomitant participant);
(59) a.
b.
c.
d.
e.
There is a situation S with its set of central participants.
There is an additional participant C whose nature may vary.
C is peripheral to S.
C‘s participation in S is in some way oriented towards some central participant.
The relation of C to S may be captured by an additional predicate; ultimately, C
may be in a situation that is ‗co-present‘.
The features that are of direct interest here are (a-d). According to (a-d) the proposition in (58),
which is the alternative of (57), is analysed as follows. John is the central participant in the
situation S of finding the way to the station. Bill is the participant C; he is peripheral to the
situation S and his participation is oriented towards John (i.e. by helping John). This largely
asymmetric relation holding between the two participants of the alternative value of the alone
reading is in direct parallel to the relation holding between the entities constituting the alternative
value of the in person reading. We conclude that the alone reading requires its associate DP to be
central in discourse, on a par with the in person reading. Their only difference lies in the way
that their associate DPs are rendered as central in prior discourse. Whereas the in person
reading‘s associate DP can be central through historical events, kinship relations, distance etc,
UCLWPL 2011.01
103
the alone reading‘s associate DP becomes central with respect to the situation denoted by its
alternative proposition (see Siemund 2000 for a similar conclusion).
Taking into consideration the overall discussion regarding the in person and alone
readings, we may safely conclude that the different interpretation between them is due to the
nature of the interacting alternative values, which in turn define the intensified entity as central in
a different way. Whereas the in person reading‘s associate DP interacts with an alternative with
the type of the brother of x (i.e. John‘s brother) (the possessive construction here is used as a
representative of the kind of alternatives that contain an entity which is defined in terms of the
entity that is intensified), the alone reading‘s associate DP interacts with one of the type x with y
(i.e. John with Bill). However, we should not confuse the effect of context on the nature of the
alternative values with their initial state in the semantics. As pointed out in the case of the in
person reading, when DEP is applied onto x, the only thing that we get is an entity dependent on
x, which becomes the brother of x, or the chair near x, or Tahiti only after its interaction with
context. We do not see any reason why we could not suggest the exact same thing for the alone
reading. This would render the x with y type of alternatives to also become contextual
instantiations of the an entity dependent on x general form of alternatives. After all, this is
consistent with the stand we take in this paper that it is the context surrounding the intensifier
that decides its interpretation. When the context provides an alternative of the type the brother of
x, then we will get the in person reading. When an alternative of the type x with y is provided,
then we will get the alone reading. More concretely, we propose that the alone reading also
denotes ID (x), which invariably contrasts with DEP (x), on a par with the in person reading.
This means that at the level of semantics, the alternatives of both readings look the same.
In order to show this whole process for the alone reading, let us assume that the alternative
value of the intensified referent in (57), John, is John with Bill. This alternative value comes
about when DEP takes John as its input and eventually (after the effect of context) maps it onto
John with Bill22, the alternative value. This process is exemplified below in (60).
(60) 〚[John] with Bill〛 = DEP (〚John〛) = 〚an entity dependent on John〛 = 〚John
with Bill〛
Following this discussion, we assume the following.
(61) Hypothesis 2: The alone reading of the intensifier is invariably an instance of
contrastive focus and always makes reference to dependent function(s).23
22
This way of presenting the alternative value of the alone reading and its associate DP may lead the reader to
the understanding that the concomitant participant forms a constituent with the subject of the sentence. Of course
this cannot be the case, as all of our examples show. We consider that the prepositional phrase containing the
concomitant participant (i.e. with y) is introduced in a VP modifying position. We will continue to refer to this
alternative value in this way, disregarding the fact that it consists of a subject and a VP modifier.
23
Similarly to hypothesis 1, hypothesis 2 states that it is the intensifier itself and not the constituent it interacts
with that is contrastively focused. The intensified constituent is interpreted as if it is contrastively focused itself
through its interaction with the intensifier. The reader should assume that this clarification applies to all the
hypotheses that follow.
UCLWPL 2011.01
104
Once we assume (61), everything falls into place with regard to the meaning contribution of the
alone reading. The notion of contrastive focus will give us the requirement of this reading to
interact with contextually given alternative propositions, or evoke them if not present, and at the
same time to exclude one of them. The reference to the dependent function will deliver the
intuitively right type of alternatives. The example below illustrates that the alone reading is
felicitous in a corrective context, characteristically suitable for contrastively focused
constituents.
(62) Speaker A: John wrote this book with Bill.
Speaker B: No, you are wrong, John wrote this book HIMSELF.
In contrast to (62), which provides the suitable alternative value (John with Bill) for the
intensified referent to contrast with, the discourse in (63) fails to do so, at least in the absence of
any accommodation on the part of the hearer, hence the infelicitous use of the intensifier.
(63) Speaker A: Who wrote this poem?
Speaker B: #William Wordsworth wrote this poem HIMSELF.
(63B) is infelicitous as an answer to (63A) because there are no contextually given alternatives
for the intensifier and its associate DP to make reference to. In actual discourse, speaker B‘s
reply would trigger a request for clarification, similarly to the case of the in person reading,
along the lines of What do you mean? Who didn’t take part in the writing?. This fact further
substantiates the claim that the alone reading requires some other referent(s) available in
discourse for the intensified referent to oppose with. (64) provides the representation of the
relevant part of (62B)24.
(64) a. <λx [x wrote this book], John, {John, DEP (John)}>
b. x [x ∈ {John, DEP (John)} &  [x wrote this book]]
(For a natural language paraphrase of the type of expressions in (64), see the discussion
surrounding (48)). What (64) essentially expresses is that the proposition in (62B) is true and the
contrasting alternative one in (62A) is false. It also expresses that the presence of the alone
reading in (62B) constrains the members of the set of alternatives in such a way that only certain
variants of the proposition John wrote this book can be included. The position held by John in
these variants can only be filled by constituents denoting referents peripheral to John, of the type
John with y, and nothing else.
The analysis presented here also explains a phenomenon only related to the alone
reading, touched upon in section 2, namely its exceptional dependence on the denotation of the
verb. As pointed out, this reading is compatible with verbs denoting some sort of
accomplishment, achievement or activity, and incompatible with stative verbs, as shown in (11),
repeated below as (65).
24
As pointed out in the previous section, this representation disregards the position of the intensifier. It is taken for
granted that the intensifier takes scope over the entire proposition, something which is not reflected in the surfaced
syntactic position of the intensifier in the case of the alone reading. As pointed out in a previous footnote, further
research is needed to determine whether the scope of the intensifier deviates in certain situations from its surfaced
position (which is essentially what is assumed here), and if this is indeed the case, under which circumstances this
process is disrupted.
UCLWPL 2011.01
105
(65) #I am a gardener MYSELF. (intended intensifier reading: alone)
As pointed out above, the denotation of the alone reading, ID (x), contrasts with DEP (x) which,
after the interaction with the context, delivers a set of alternative values of the type x with y.
These alternative values replace the intensified referent x in the respective alternative
propositions. What matters here is the fact that the evoked alternative values consist of the
comitative construction, the prepositional phrase with y. Similarly to (57), in which the presence
of the alone reading evokes alternative propositions of the sort John found the way to the station
with Bill, (65) also evokes alternative propositions of the sort in (66) (the referent Bill is just an
instantiation of y). However, such a proposition is ungrammatical.
(66) *I am a gardener with Bill.
Stative verbs are intrinsically incompatible with the meaning of the comitative PP with y. This
PP is only compatible with verbs denoting some sort of action, similarly to the alone reading that
it contrasts with. Therefore, we propose that the incompatibility of the alone reading with stative
verbs is a consequence of the nature of alternative propositions that it interacts with (which also
partly determine its interpretation). Since there are no grammatical alternative propositions
consisting of both a stative verb and the PP with y (that is consistently evoked with this reading),
there cannot exist alternative propositions at all, hence the incompatibility of the alone reading
with stative verbs.
The assumption that the alternative values of the referent intensified with the alone reading
are structured as x with y, sheds light on one other fact, again touched upon in section 2, namely
the intensifier‘s requirement to occur post-verbally. Again, the nature of the evoked alternative
propositions seems to be a decisive factor with respect to the intra-propositional distribution of
the alone reading. Consider the examples below.
(67) *John, with Bill, found the way to the station.
(68) John found the way to the station with Bill.
According to our analysis, both (67) and (68) can, in principle, be alternative propositions to the
sentence in (57). However, only (68) is grammatical due to the requirement that part of the
alternative value of John, the comitative PP with Bill, must occur post-verbally. This is because
this PP acts like a VP modifier, and in English VP modifiers must occur after the verb. Now, it is
well known that there is a general constraint governing the structure of contrasting propositions
and utterances. That is, the contrasting elements must have the same number and type of major
constituents. Let us call this constraint the parallel (major) constituency constraint (PCC).
According to Hankamer (1973: 18), ―A "major constituent" of a given sentence S0 is a
constituent either immediately dominated by S0 or immediately dominated by VP, which is
immediately dominated by S0.‖ In other words, complete subjects, objects and adverbials qualify
as major constituents, but not fragments of any of them (Ning Zhang 2007). In order to illustrate
how the PCC works, take the sentence below, in which the secondary predicate drunk (another
major constituent) is marked as contrastive focus.
(69) John entered the room DRUNK.
UCLWPL 2011.01
106
The proposition in (69) is understood to contrast with something like (70) (the choice of the
secondary predicate sober is accidental).
(70) John entered the room sober.
It cannot be the case that (69) contrasts with something like John entered the room, in which
there is no secondary predicate, despite the fact that it is perfectly grammatical, because the
contrasting propositions must consist of the same number and type of major constituents. This
point may seem too simple, or obvious to state, but it is enough to explain the post-verbal
character of the intensifier on the alone reading, as well as its free distribution on the in person
reading. Consider the structure of (68). This example can be analyzed as consisting of three
major constituents (various constituency tests can confirm this, i.e. do so substitution) namely the
DP John (subject), the VP found the way to the station, and the VP modifier with Bill. According
to the PCC, the contrasting proposition of (68) must consist of the same number and type of
major constituents. Assuming that (57) is this contrasting proposition, the only constituent found
in this proposition that has the general ability to merge in various positions is the intensifier (see
section 2). However, if it merges anywhere except in a VP-modifying position, which is a major
constituent position, the PCC is violated (because one of them will have a VP-modifier and one
will not). Therefore, the intensifier must serve the role of a VP-modifier (apart from its
intensifying role). Since VP-modifiers are always post-verbal in English, the alone reading is
consistently found in that position. This conclusion is corroborated by the cross-linguistic data.
There are languages in which the comitative PP with y is not found post-verbally. In Dutch for
example, such phrases can merge pre-verbally. In fact, they can merge in various positions, as
shown in the example below in which the with y phrase can merge either to the left or to the right
of the direct object depending on whether the direct object is old (first bracket from the left) or
new (second bracket from the left) information.
(71) Jan heeft (het huis) met Marie (een huis)
John has (the house)with Mary (a house)
John has built a/the house with Mary.
gebouwd.
built.
According to our analysis, this predicts that the alone reading of the Dutch intensifier should
have a similar distribution. The example below shows that zelf follows a similar distribution to
the with y phrase. Note that we put in brackets all the possible positions that zelf can merge.
(72) Jan
(*zelf)
heeft
(zelf)
Jan
(*zelf)
heeft
(*zelf)
John has built a/the house HIMSELF.
een huis
het huis
(*zelf)
(zelf)
gebouwd.
gebouwd.
Being a VP modifier, the alone reading is predicted to pass all the tests that show its containment
in the VP. Indeed, Gast (2006, ch.4) shows that (only) this reading can form part of deverbal
nouns and adjectives, can be left-dislocated together with a VP (see also Browning 1993) and is
always c-commanded by, and in the scope of, VP-external adverbials.
The freer distribution of the in person reading (see section 2) now falls out naturally.
Remember that the alternative propositions of the proposition containing the in person reading
look different from the ones contrasting with a proposition containing the alone one. According
to the analysis provided in section 3.3. for the in person reading, (73a-b) below can be two
UCLWPL 2011.01
107
contrasting propositions (the intensifier in brackets shows all the possible positions that it can
occur). They have the exact same major constituency structure, namely a DP acting like the
subject (the director/the director‘s secretary), and a VP consisting of the verb (has appeared) and
a PP acting like the direct object (to the meeting). What this means is that the PCC is satisfied
without the presence of the intensifier. Therefore, there is nothing blocking the intensifier from
merging in any of the positions shown in (73a). Note that this argument goes through only if we
adopt the widely accepted view that the in person reading is a DP modifier (see Siemund 2000,
Bergeton 2004 among others), thus not a major constituent, which may merge in various
positions in the sentence (something like a floating quantifier). Thus, according to the PCC the
two propositions below are still parallel, despite the presence of the intensifier in (73a).
(73) a. The director (HIMSELF) has (HIMSELF) appeared to the meeting (HIMSELF).
b. The director‘s secretary has appeared to the meeting.
This analysis predicts that the presence of himself when it carries the in person reading is not
obligatory, whereas it is on the alone reading. The examples below show that this prediction is
borne out. As illustrated in (74), the DP interacting with the intensifier in (73a) can be marked as
contrastive focus, and not be intensified, without loss of felicity. Conversely, marking the DP
interacting with the alone reading in (57) (only) as contrastive focus sounds odd, as it leaves us
with the impression that something is missing, as illustrated in (75).
(74) Speaker A: The director‘s secretary has appeared to the meeting
Speaker B: No, THE DIRECTOR has appeared to the meeting
(75) Speaker A: John found the way to the station with Bill.
Speaker B: ??No, JOHN found the way to the station.
Note that (75B) is not completely infelicitous due to pragmatic accommodation on behalf of the
hearer. This accommodation runs along the following lines of general Gricean reasoning; a)
speaker B negates the proposition in which both John and Bill participate in the action denoted
by the predicate, b) speaker B simultaneously asserts that John participates in that same action,
c) speaker B is assumed to be maximally informative, therefore it must be the case that Bill did
not participate in the action denoted by the predicate because otherwise the speaker would have
said so. This conclusion equals to the negation of the version of the event of finding the station in
which both John and Bill participate. In other words, there is negation of the proposition uttered
by speaker A through pragmatics, something which leads to a similar interpretation of (75B) and
(57).
We now move on to discuss the exhaustive interpretation of the alone reading. As noted in
footnote 19, the in person reading behaves as predicted by the semantics of contrast, which
expresses that not all alternative propositions are excluded by its use. As the example below
shows however, the alone reading behaves somewhat differently, in the sense that it excludes all
alternative propositions in discourse. This poses an obvious problem for our semantic analysis in
(64), which consists of the semantics of contrast. Remember that the semantics of contrast
expresses through the use of the existential operator, which is in the scope of the negative
operator, that at least one (and not all) alternative proposition is negated. Until now we got away
with the use of the existential operator because we would always have examples with one
alternative proposition. What would do the job then, is the replacement of the existential with a
UCLWPL 2011.01
108
universal operator. In this way, all alternatives within its scope would be negated. Despite the
attractive simplicity of this proposal, it raises an important conceptual issue. This change at the
level of semantics is ad-hoc. It is not predicted by the theory of contrast, as outlined in section
3.2. This extension is made up and stipulated specifically for the sake of the exhaustive
interpretation of this reading. As it becomes apparent, assuming different semantics for the alone
reading would be an undesirable outcome for our attempt to account for all the readings of the
intensifier using the system of notions depicted in (27). Fortunately, for our theory, there is no
need to make such extra assumptions. As explained below, independent reasons and the
assumptions made up to now suffice to explain the (apparent) exhaustive character of the alone
reading.
(76) Speaker A: John found the way to the station with Bill.
Speaker B: No, John found the way to the station WITH MARY. (*Furthermore, he found
it with George)
Speaker C: No, John found the way to the station WITH PETER. (*Furthermore, he found
it with Bill)
Speaker D: No, John found the way to the station HIMSELF. . (*Furthermore, he found it
with Mary)
As shown in (76D), the use of the alone reading is indeed exhaustive. Speaker D negates all of
the alternative propositions being salient in discourse, as it cannot be licitly continued by one of
its alternative propositions. However, it can be seen that this reading shares this characteristic
with the rest of the constituents marked as contrastive focus in (76). None of the propositions in
(76) can be continued with an alternative proposition found in discourse, or, as shown in (76B),
with a proposition which is not salient in discourse. As can be concluded from (77), this cannot
be due to the presence of the negation. (77), which is an out of the blue proposition, also shows
that the marking of the VP modifiers as contrastive focus is irrelevant to the phenomenon shown
in (76)25.
(77) John found the way to the station with Bill. *Furthermore, he found it with Andrew.
It is important to keep in mind that the intended interpretation of the sentences in (76) and (77) is
not of the type John found the way to the station with Bill and Andrew. Instead, the intended
interpretation is that there is an event of finding the way to the station. Depending on the referent
denoted by the comitative PP, the version of that same event changes (to be explained below). It
follows that one and the same speaker cannot state more than one truthful version of the same
event. It would be a contradiction. Why would this happen when himself is used on the alone
reading (or with the rest of the VP modifiers) and not when it is used on the in person reading?
We believe that this follows from the fact that the intensifier on its alone reading, but not on its
in person reading, is a VP modifier. As indicated by their name, VP modifiers modify the VP.
Since the VP is the constituent denoting the type of event denoted by the whole proposition, VP
modifiers take that event and change (or modify) it accordingly at the level of interpretation.
Therefore, one cannot utter more than one sentence with a comitative VP modifier (which is
25
At this point, the reader may be left wondering as to what is the empirical basis for assuming that the alone
reading of the intensifier is marked contrastively. The main empirical argument for this assumption is given in
section 3.8.
UCLWPL 2011.01
109
what the intensifier on the alone reading is26 and contrasts with) because that would amount to
stating more than one version of the same event. A similar restriction is not found with the
intensifier on its in person reading, as it only interacts with a DP. Since a speaker can modify a
DP more than one time in one or more utterances denoting the same event, without resulting in a
contradiction (i.e. John‘s brother and John‘s sister found the way to the station. John‘s mother
helped them to find the way to the station, too.), the in person reading has no problem in
exhibiting its contrastive (and not exhaustive) character. This is because DP modifications
simply result in different referents taking part in the same event. This means that the in person
reading has the choice of excluding, as well as including, some of them (see footnote 19 for an
example). On the other hand, a sentence containing the intensifier on the alone reading is
incompatible with the truth of some of the alternatives, because a true alternative proposition
would contradict the proposition containing the alone reading itself. We conclude that the
exhaustive character of the alone reading is only apparent and falls out from independent
reasons, namely its syntactic position (VP modifier) within the sentence. It follows that there is
no need to change the semantics of this reading in (64).
It has been shown that the (apparent) exhaustive interpretation of the alone reading is due
to its syntactic status, and not its semantics. This claim is based on the fact that every VP
modifier of the comitative type (including the alone reading) exhibits the same behaviour.
However, the example below shows that the alone reading exhibits something unique when
compared to the rest of the VP modifiers of the comitative type.
(78) Speaker A: John found the way to the station with Bill and (with) Sue.
Speaker B: No, John found the way to the station HIMSELF (*AND (WITH) MARY).
As illustrated in (78), a comitative type VP modifier can consist of more than one referent. Since
the alone reading is assumed to be a VP modifier of the same type, we would expect it to be able
to behave in the same way. This is not the case however, as shown in (78B). We think that the
answer to this problem lies in one of the core functions of the intensifier. As noted above, the
alone reading requires a value that is central to interact with. Let us see whether this requirement
is fulfilled in (78). The alternative proposition of (78B), in (78A), consists of the alternative
value John with Bill and Sue. This alternative value comes about when DEP (= with Bill and Sue
when contextualised) is applied onto John. Now, the function denoted by himself, ID, is applied
onto John, in (78B), and delivers John. However, this time the outcome of this process is
coordinated with Mary to deliver John and Mary. As can be seen from a comparison of the two
alternative values, John and Mary is not central against John with Bill and Sue. This is because
Mary is not even present in the alternative value to be able to be defined as central. Hence, the
impossibility of combining himself and and Mary in a context like (78). The example below
illustrates however that even if Mary is present in the alternative value, it is not sufficient to be
rendered as central.
(79) Speaker A: John found the way to the station with Bill and (with) Mary.
Speaker B: No, John found the way to the station HIMSELF (*AND (WITH) MARY).
26
As a side note, the assumption that the alone reading is a VP modifier that behaves, at least syntactically, in
the same way as the comitative PP with y also explains its subject orientation, as it is well known that the comitative
with y PP only interacts with subjects, and never objects. This is shown below.
a)
John1 met Mary2 [PP with Jim1/*2].
UCLWPL 2011.01
110
The resulting contrasting values of the example above are John with Bill and Mary and John and
Mary. Both referents of the latter value are present in the former one, thereby no issue arises with
regard to the entities being present in the alternative value. However another issue arises.
Consider the way the intensified value is being formed in (78) and (79). ID is applied to John to
give us John. Only after this process we get John and Mary. However, the only referent of the
two that has previously satisfied ID is John. Therefore, the resulting value, John and Mary, is
odd in the sense that half of it requires to have been defined as central (because it has satisfied
ID) in the alternative value, and half of it does not. This is impossible of course. The only way
for John and Mary to satisfy ID as a whole is when both referents form one DP (for syntactic
reasons, see Constantinou 2011), as shown below. Thereby the possibility of the second
conjunct, and Mary, to be found in any other position in the sentence is ruled out.
(80) Speaker A: John and Mary found the way to the station with Bill.
Speaker B: No, John and Mary found the way to the station themselves.
As noted at the beginning of this subsection, the use of the alone reading can give rise to
the paraphrase without x’s help. In fact, for the example in (62B), repeated below as (81), such a
paraphrase would perhaps be more appropriate than one using alone.
(81) John wrote this book HIMSELF.
Abstracting away from the paraphrase possibilities of the above example (which can be
misleading), what (81) certainly does is to imply that John wrote the book without someone’s
help. The identity of the helper is specified through contextual or world knowledge clues,
similarly to any other reading of the intensifier we have seen so far. Even though this reading has
led various researchers to assume a separate lexical entry (see, for example, Eckardt 2001), we
believe that it is merely another instance of the intensifier being interpreted as a contrastive focus
whose evoked alternatives result from the application of the dependent function, which in turn
yields (after the effect of context) the type of alternative x with y (comitative) . The refinement to
‗without someone‘s help‘ is a pragmatic effect resulting from the kind of context the intensifier
occurs in. Similarly, the alone paraphrase can also be viewed as a pragmatic enrichment of the
basic opposition between x and x with y. The type of event denoted by the proposition under
discussion is one factor determining the nature of context. In (81) for instance, the proposition
denotes an agent (John) participating in the action of writing a certain book. At the same time
(81) asserts that this agent did not perform this action together with some contextually given
person. The negation of this alternative proposition implies that the referent y did not participate
in the action of writing this book. Due to the nature of the event, this lack of involvement or
contribution of y is (usually) paraphrased with without y’s help. It cannot imply (thus be
paraphrased with) something of the sort not accompanied by y, as in the case of (82) below,
because it does not make sense in that particular context.
(82) John went to the station HIMSELF.
UCLWPL 2011.01
111
It is inconceivable for someone to be accompanied, in the literal sense of the word, in the writing
of a book. What can be concluded here is that the alone and without someone’s help27 readings
are all instances of the same occurrence of the intensifier. These paraphrases are a result of
pragmatic enrichment of the basic meaning of the intensifier, which we have expressed in terms
of the function that generates alternatives, which are in turn realised in the specific comitative
type x with y.
3.5. Interim summary
We began this section with the assumption that the intensifier denotes the identity function ID
(x), which takes as its input a nominal constituent x and delivers the (same) output x (Eckardt
2001). Then we reviewed Neeleman and Van de Koot‘s (2008) typology of information
structural categories and hypothesised that the intensifier‘s various readings partly depend on the
nature of its information structural marking. We have demonstrated that the in person and alone
readings, along with their associate DPs, are interpreted as such because of their marking as
contrastive focus and the contextual realisation of the alternative to ID, the dependent function
(DEP). Whereas the in person reading contrasts with alternatives of the type brother of x or the
chair near x or simply Tahiti, the alone reading contrasts with alternatives of the (comitative)
type x with y. Both types are instantiations of an asymmetric relation (expressed by DEP (x))
holding between two entities. Unless these instantiations are ruled out for independent reasons,
we expect them to interact with the intensifier, and influence its interpretation, in instances in
which the intensifier is marked with the rest of the information structural categories of (27). On
the other hand, we do not expect to find contextual instantiations of DEP (x) which do not
exhibit an asymmetric relation between the entities constituting the alternative values. This last
prediction is discussed below, in section 3.6.
3.6. The (alleged) only reading and the idiosyncrasies of each instantiation of DEP (x)
Before discussing in detail the only reading, we clarify (once again) that the labeling of each
reading is just one way of stating what type of alternative value each reading interacts with and
how it is information-structurally marked28. Therefore, the labeling only is our terminological
way of saying that the alternatives of a contrastively focused intensifier are of a coordination
type. As will become clear, the point of the first half of this section is to show that the intensifier
cannot interact with alternatives whose constituent entities do not relate asymmetrically. The
second half provides an argument in favour of the view that it is the asymmetric relation holding
between the constituent entities of the alternative that is saliently evoked when the intensifier is
contrastively marked and not alternative referents per se.
In section 2, we presented example (14), repeated below as (83), and claimed that the
intensifier could potentially be interpreted similarly to the focus particle only.
(83) Speaker A: Mary gave Jane her syntax course-book and all the notes before leaving the
university. At least that what she told me!
27
For both reasons of space and to avoid repetition, we do not provide a fully fledged analysis of this reading
along the lines of the alone one. The reader should assume that the analysis presented above for the alone reading
also applies to the without someone’s help one.
28
As we will see later in this paper, this is not entirely true for the case of the also reading as there is an
additional factor determining its interpretation.
UCLWPL 2011.01
112
Speaker B: ?Well, you must have misheard! Mary gave the syntax course-book ITSELF,
and threw away the notes. (paraphrase: ... Mary gave John only the syntax ...)
The reason for being led to such a conclusion is the interpretation of the sentence in (83B)
by English native speakers as excluding the notes from the event of giving. However, English
speakers also point out that the sentence is not completely felicitous. Note that the relevant
interpretation is possible only when the intensifier is marked with an A-accent, characteristic of
contrastive focus.
In view of the above, it could be argued that the only reading is just another instance of the
alone reading. Both readings and their associate DPs are interpreted as contrasting with a type of
alternative whose constituent entities denote a separate referent in the proposition (compare
John’s brother with John and Bill/John with Bill). Moreover, both instances exclude the referent
of the alternative which is not subsequently intensified. However, the analogy ends here. The
only reading‘s alternative is a conjunct whereas the alone reading‘s is a VP modifier. More
importantly, the felicity judgements do not allow us to pursue an analysis along the lines of the
alone reading. The judgements suggest that there must be something else going on here.
In fact our analysis predicts that the intensifier should not be able to interact with types of
alternatives which are formed by coordination. This is because ID (x) requires that there is an
asymmetric relation between the entities constituting the alternative value. Coordination does
precisely the opposite. It imposes a symmetric relation between the coordinated entities. Hence,
the only reading should not exist in the first place.
We think that this is correct and the mild infelicity of (83B) is an indication of this. The
reason that (83B) is not completely infelicitous is because of the entities that comprise the
alternative value. The notes are interpreted as being peripheral to the syntax course-book. This
asymmetric relation between the two entities mitigates the symmetric relation that the
coordination imposes on them. Once the coordinate entities in the alternative value become of
equal status, the felicity judgements become much worse compared to (83B), as shown below in
(84).
(84) Speaker A: John and Bill have become incredible students lately.
Speaker B: #No, John HIMSELF has become an incredible student lately.
Since the intensifier is not responsible for the only inference in (83), the question arises as to
where this inference comes from. We think that the only inference is a result of general Gricean
reasoning which runs along the following lines; a) speaker B negates the proposition in which
both John and Bill participate in the event denoted by the predicate, by marking the intensifier as
contrastive focus b) speaker B simultaneously asserts that John participates in that event, c)
speaker B is assumed to be maximally informative, therefore it must be the case that Bill did not
participate in the event denoted by the predicate because otherwise the speaker would have said
so. This conclusion amounts to the only inference. Hence, factors independent from the presence
of the intensifier are held responsible for it. The example below provides further support to this
conclusion. As can be seen, the only inference is still present without the use of the intensifier.
Our explanation for the existence of this inference when the intensifier is present (as in (83B))
applies to the example below too.
UCLWPL 2011.01
113
(85) Speaker A: John and Bill have become incredible students.
Speaker B: No, JOHN has become an incredible student.
The purpose of the discussion of this section up to this point is two-fold. One, to illustrate that
the only reading does not exist (and hence neither a coordination type of alternatives). This has
provided further support to our assumption that the alternative value of the denotation of the
intensifier, ID (x), is DEP (x). Two, to argue against the view that the centrality effect found in
intensifier constructions is a result of the presence of the intensified value in the alternative
values (see Eckardt 2001, Hole 2002). If this were the case, then the non-existence of the
coordination type of alternatives (and hence the only reading) remains unexplained, unless
further assumptions are made (see Gast 2006 for an account which restricts the alternatives using
focus suppositions (a weak type of presuppositions prior to the moment of the utterance)).
In the remainder of this section, we provide one other argument in favour of the existence
of the DEP (x) function, and its interaction with ID (x). In doing so, we also further support the
distinction between contrastively and non-contrastively marked readings of the intensifier.
Finally, we discover how the linguistic constructions instantiating DEP (x) differ with regards to
their specification of (un)ambiguous relations between the entities comprising them and by
extension how this affects the alternatives of each reading of the intensifier.
The following asymmetry between the alternatives of the alone reading on the one hand
and the alternatives of the in person reading on the other appears to hold.
(86) Speaker A: John found the way to the station with someone.
Speaker B: No, John has found the way to the station HIMSELF .
(87) Speaker A:*John‘s someone looks angry.
Speaker B: #No, John HIMSELF looks angry. (intended reading: in person)
As illustrated in (86), the comitative construction has no problem in partly consisting of the
indefinite pronoun someone, which is however read specifically. More importantly, the alone
reading and its associate DP are perfectly able to interact with such alternative. On the other
hand, it can be seen in (87) that the in person reading and its associate DP cannot interact with
such alternatives. At first sight this asymmetry is unexpected. One could argue that the infelicity
of (87B) is a direct result of the possessive construction‘s inability in (87A) to accommodate an
indefinite pronoun, such as someone. However, an explanation along these lines is unattainable
once we consider the example below, which is a variation of (40).
(88) Speaker A: #I have never been to Hawaii ITSELF, but I‘ve been to some (other) place.
(intended reading: in person)
In (88), Hawaii is intended to contrast with some place (see example (40) in which Hawaii
contrasts with Tahiti). Contrary to the case of (87), the contrasting alternative referent of the in
person reading and its associate DP is a perfectly grammatical construction. Despite this, the use
of the in person reading of the intensifier remains infelicitous. Things become even more
complicated when considering the example below, in which the in person reading and its
associate DP felicitously contrast with an alternative that consists of the indefinite pronoun
something. This is shown in the example below, which is a variation of (45).
UCLWPL 2011.01
114
(89) Speaker A: Something near the table must smell really awful.
Speaker B: Well, I think that the table ITSELF smells awful. It‘s been a week to clean it.
In order to put things into perspective, we outline the problem presented by a comparison of the
examples (86) – (89). The alone reading has no problem with interacting with alternatives which
are partly comprised of entities denoted by indefinite pronouns (ie. x with someone). Similarly,
the in person reading can interact with such alternatives when it comes to the DP modification
instantiation of DEP (x) (see (89)). But not when it comes to the possession instantiation (see
(87)) (in fact possession cannot accommodate indefinite pronouns) or the one entity instantiation,
such as Tahiti in (40), (see (88)) of DEP (x). In a nutshell, what seems to be relevant here is the
particular instantiation of DEP (x). Up to this point we have explicitly distinguished two types of
alternatives (instantiations of DEP (x)), namely the comitative type and all the rest (i.e.
possession, DP modification, Tahiti). This distinction was based on the different readings we get
when the intensifier is contrastively focused. In addition, there is an obvious difference with
respect to the nature of the (asymmetric) relation these two kinds of alternatives express. The
comitative type expresses a participatory relation between two entities. The choice of this
terminology is based on the features of concomitance as stated in (59). With regards to the other
type of alternatives, we can generalize and state that they express some kind of defining relation
between the entities that comprise them. This characterization is based on the intuition that the
one entity in these constructions is invariably defined in terms of the other. For instance, in
John’s book, the book is defined in terms of John; in the chair near the table the chair is defined
in terms of the table; and in Tahiti, as in example (40), Tahiti is defined in terms of Hawaii.
However, this distinction seems to be of no use to our attempt to understand what is going on in
(86) – (89) as the in person reading seems to be exhibiting an irregular behavior.
To solve this puzzle, we need to take a closer look at each of the contextual instantiations
of DEP (x), irrespectively of which reading of the intensifier they deliver. Before doing this, let
us see what our approach predicts. Remember that the in person and alone readings are
contrastively focused. This means that ID (x) evokes a salient alternative, which is DEP (x). DEP
(x) is the expression of an asymmetrical relation, which takes various contextual instantiations.
As a result, we predict that the expression of this asymmetrical relation will be salient. Having
this in mind, we first attempt a comparison between the relations expressed by the chair near the
table and John’s book. The relation holding between the the chair and the table is clearly a
relation of distance, as expressed by the preposition near (this is pretty clear-cut). On the other
hand, it seems rather difficult to define the relation between John and the book. As Nikolaeva
and Spencer (2010) point out the list of possible interpretations of the relation between the
possessor and the possessee can be extended indefinitely, subject to contextual factors. In our
example for instance, the relation between John and the book can be that of ownership,
authorship or even the fact that John is holding the book. In a nutshell, possession is essentially a
linguistic construction that is able to realize more than one relations. Since the precise definition
of the relation between the possessor and the possessee is subject to context, it follows that this
cannot be achieved without identifying the entities being at the two ends of this relation. A direct
result of this is the fact that it will never be able to accommodate an indefinite pronoun, even if it
is read specifically, thus we explain the ungrammaticality of (87A). Crucially, this way of
thinking also explains the infelicity of (87B). (87B) is infelicitous because DEP (x) in (87A) had
not even been defined for it to be salient, as required by the contrastively focused ID (x).
UCLWPL 2011.01
115
However, one could still argue that the infelicity of (87B) is due to the ungrammaticality of its
alternative. (88) provides the means to maintain our last conclusion. The alternative referent of
the in person reading and its associate DP, Hawaii, is some place. This last expression is
grammatical; nevertheless the use of the in person reading is infelicitous. The reason for this is
precisely the same as it is for the case of the possession construction. Remember that in order for
Hawaii to be able to be intensified with the in person reading and contrast with Tahiti in example
(40), there must be a contextual relation holding between the two entities (i.e. financial
dependence of Tahiti). Since we replaced Tahiti with an indefinite pronoun, some place, this
relation cannot even be defined for it to be salient, hence the infelicity of (88). It follows that in
order for the last two constructions we have discussed to be felicitous alternatives to a
contrastively marked intensifier, only identifiable entities can comprise them.
In light of the discussion of the last paragraph, we can now explain why the construction
something near the table in (89A) is grammatical as well as why the in person reading of the
intensifier in (89B) is used felicitously. As pointed out above, the preposition near, which is used
to introduce the modifying expression of the subject DP of (89A), expresses a relation of
distance. Contrary to the case of the possession construction, in the subject DP of (89A) there is
a one to one mapping between a linguistic expression and a relation. That is to say, the
preposition near invariably denotes a relation of distance (but not the other way around). A result
of this is primarily the fact that an indefinite pronoun can enter such a construction. This is
because the possibility of accommodating (on behalf of the hearer) the identity of the entity
denoted by the indefinite pronoun now arises. Once the identity of one of the entities of the near
construction and the relation holding between the constituent entities is known, the search space
of identifying the indefinite pronoun denotation is sharply reduced. In our example for instance,
the hearer knows, through the unambiguous relation denoted by near and the unambiguous entity
denoted by that the table, that the denotation of something must be a thing which lies in close
proximity to the table, hence the possibility of accommodating the thing‘s identity. On the other
hand, in the case of possession, the search space of identifying the indefinite pronoun‘s
denotation is in principle infinite (because we do not know the relation). This is an impossible
task to do, hence the disallowance of an indefinite pronoun to form part of a possession
construction. It is important to keep in mind that the indefinite pronoun, whose default
interpretation is that of a true indefinite, is forced to be read specifically because it participates in
a relation. This is because a true indefinite denotes a group of properties and a specific indefinite
denotes an individual, and only individuals can participate in a relation29. The possibility of using
the in person reading in (89B) now falls out naturally. ID (x) is contrastively marked, hence
saliently evokes DEP (x). DEP (x) is successfully realized by the asymmetrical relation holding
between something and the table, as it is clearly defined in the salient alternative proposition
through the use of the preposition near. Hence, the felicity of (89B).
The exact same reasoning applies to (86), which we summarize below. The possibility of a
comitative construction to accommodate an indefinite pronoun (86A) results from the fact that
this linguistic expression always denotes a participatory relation (as defined in (59)). There is no
ambiguity, hence it can accommodate an indefinite pronoun, which is read specifically for the
29
We clarify that the possession construction can force an indefinite pronoun to be read specifically, as
intuitively understood. This is because there is a relation (albeit undefined) holding between the entities, as
expressed by ‘s in John’s something. As explained above, the problem lies in the fact that neither the (specifically
read) indefinite pronoun denotes an identifiable entity nor the relation denoted by possession is unambiguous.
UCLWPL 2011.01
116
reasons outlined in the previous paragraph. A direct result of its non-ambiguous status is the fact
that the saliently evoked DEP (x) is successfully realized in context, and thus the felicitous use of
the alone reading of the intensifier in (86B).
The second half of this section has provided further support to the assumption that the
alternative of the intensifier is a relation, as denoted by DEP. The presence of individuals in the
alternatives is a result of the inducing of this relation30. In addition, we have explained why
certain linguistic constructions allow for an indefinite pronoun to form part of them. Even though
we have not given counter-examples, we advocated that the reason behind the (in)ability of the in
person and alone readings of the intensifier to interact with certain constructions partly
consisting of indefinite pronouns is the notion of contrast. The prediction is that a noncontrastively marked intensifier will be able to interact with such alternatives, once these
alternatives themselves allow for such configuration (see for instance (88)). As we will see later
on, this expectation is borne out.
3.7. The as for reading
This reading has largely remained unexplored throughout the literature. Eckardt (2001) and Féry
(2010) are the only researchers, to our knowledge, mentioning this potential use of the intensifier
without, however, going into great detail with regard to its meaning contribution to the sentence.
(90) exemplifies its use (remember that double underlining indicates a B-accent; in English,
switch topics are invariably marked with this accent).
(90) Speaker A: Tell me about John‘s brother. How is he taking the whole moving to Paris
issue?
Speaker B: Well, I don‘t know about John‘s brother, but John himself is fine with it.
When using the as for reading in (90B), the speaker is understood to be performing the following
speech acts: a) Consider John out of the set of possible alternatives John and John’s brother; b) I
assert that John is fine with the moving to Paris issue. To put it differently, the utterance in (90B)
asserts something about the subject referent, suggesting that the subject is a topic. In terms of its
discourse function, the question in (90A) with the general form ―Tell me about X‖ invites
speaker B to make X (John‘s brother) the topic of his/her utterance and say something about this
X. Speaker B‘s answer however, switches the topic from X to Y (John), something which is
achieved with the use of the intensifier. Notice that the intensified referent is understood to
interact with another referent explicitly mentioned in discourse, namely John’s brother. This is
reminiscent of the interpretation of the previous readings of the intensifier we have seen, which
invariably (force the intensified referent to) make reference to alternatives present in discourse.
We argued that this is a consequence of those readings being specified with contrast. The
question then is whether this instance of the intensifier also makes reference to such alternatives
obligatorily, or whether the reference to a specified set of alternatives, as in (90B), is merely a
consequence of the context in (90). (91) gives us a first indication regarding this.
30
a)
Note that it is possible for non-intensified indefinite DPs to contrast, as shown below.
Speaker A: This is a dog.
Speaker B: No, A WOLF this is.
Importantly, the two indefinites are not read specifically, but are true indefinites. This further substantiates our claim
that ID (x) interacts with DEP (x).
UCLWPL 2011.01
117
(91) John himself gave the presentation he was supposed to.
(91) forces us to consider someone that could potentially take the place of John in performing the
action described by the sentence. In fact, (91) is infelicitous if uttered out of the blue (and there is
no accommodation on behalf of the hearer). This is because the shared world knowledge
between the interlocutors does not specify anyone apart from John about whom an assertion that
he did or did not give the presentation he was supposed to could in principle be made. In such a
situation, uttering (91) would most likely trigger a request for clarification, such as What do you
mean?, in an attempt on the part of the hearer to understand who else should be under
consideration. The fact that this reading makes reference to explicit alternatives in discourse
obligatorily is confirmed through a comparison of (92) and (93).
(92) Speaker A: Tell me about your friend Maxine.
Speaker B: Well, [Maxine]T was invited to a party by Claire on her first trip to New
York.
(93) Speaker A: Tell me about your friend Maxine.
Speaker B: #Well, Maxine herself was invited to a party by Claire on her first trip to
New York.
The use of a topical constituent in (92B) is appropriate in response to the request from speaker.
A. Marking Maxine as a topic serves the function of introducing her as the topic of discourse,
following speaker A‘s instruction to do so. This topic is selected out of an undefined set of
possible topics, which in principle can be infinite. It is this undefined nature of the set of
alternative topics that makes (92B) felicitous. In discourse terms, the use of such a topic does not
require the interlocutors to have alternative topics under consideration. The use of the as for
reading in (93B) also introduces Maxine as the topic of discourse. Crucially though, the B-accent
on the intensifier expresses a discourse move constituting a switch of the topic31 that speaker B is
previously invited to say something about, something which does not happen, and a
simultaneous contrast between this ‗new‘ topic and the one speaker B is invited to comment on.
The ‗new‘ topic must be selected out of a contextually specified set of explicit alternatives that
could potentially replace Maxine. This set is not provided in the context of (93) (the set consists
of only one member, hence there cannot be contrast), hence the infelicitous status of the
utterance (in the absence of an appropriate accommodation on the part of the hearer). A similar
case is illustrated below in (94), in which – in the absence of accommodation on the part of A –
the use of a contrastive topic constituent is infelicitous because the context does not make
available any appropriate alternatives for the female popstars to contrast with.
31
Note that contrastive topics are not restricted only to switch topics. A narrow down topic is also an instance
of contrastive topic as illustrated in the by now classic example below.
a)
Speaker A: What did the popstars wear?
Speaker B: Well, the female popstars wore red caftans.
In the example above the female popstars is understood to contrast with the male popstars. Both values constitute
part of the set of popstars. As expected, the as for reading of the intensifier has no problem in performing the
function of a narrow down topic, as shown below.
b) Speaker A: How did the employees find the excursion to the mountain?
Speaker B: Well, the director himself hated it.
UCLWPL 2011.01
118
(94) Speaker A: Tell me about the female popstars.
Speaker B: #Well, the female popstars, gave an incredible performance.
The shared characteristics between contrastive topics and the kind of intensified DP found in
(90B) extend to their interpretative contribution in the sentence (at least to some extent).
Remember that the interpretational effect associated with contrastive topics is that the speaker is
unwilling or unable to make (at least) one alternative assertion. In the same vein, the use of the
as for reading in (90B) is understood as the speaker being unwilling to utter an alternative of the
sort John’s brother is fine with the moving to Paris issue because, as stated, he/she does not
possess that knowledge. However, what differentiates the meaning contribution of a contrastive
topic from that of an intensified constituent with the as for reading is the position held by the
referents of these constituents in the constructed set. On a par with the previous intensifier
readings, the as for reading requires the referent of the intensified head-DP to be central. As
discussed in sections 3.3 - 3.6, the understanding of the intensified entity as central is a result of
the nature of the entities found in the alternative value. In particular, the as for reading requires
the alternative referents replacing the intensified referent in the alternatives to be defined in
terms of the referent of the intensified DP, in the same fashion as with the in person reading (a
defining relation must hold between the two entities). Due to this requirement, (95B) is an
infelicitous answer to (95A) (compare with (90)), assuming that there is no contextual relation
between the two referents that places Bill in the entourage of John.
(95) Speaker A: Tell me about Bill. How is he?
Speaker B: #Well, I don‘t know about Bill, but John himself is fine.
As evident in (96), a contrastive topic constituent (in (96B)) does not impose such restrictions on
the referents of the constructed set.
(96) Speaker A: Tell me about Bill. How is he?
Speaker B: Well, I don‘t know about Bill, but John is fine.
In line with the above observations, we suggest that this instance of the intensifier is interpreted
as a contrastive topic. The centrality effect it imposes on the referent of its associate DP is the
result of the function it denotes, ID (x), being contrasted with the dependent function (DEP (x)).
As discussed for the in person reading, the application of this type of function onto an entity x
renders x central with respect to another entity y. The hypothesis in (97) captures the underlying
mechanisms held responsible for the meaning contribution of this reading.
(97) Hypothesis 3: The as for reading of the intensifier is invariably an instance of a
contrastive topic and always makes reference to dependent function(s).
We may now formalize the interpretation of (90B) as in (98), where the contribution of the
intensifier is restricted to the functor that generates alternatives to John.
(98) a. <λx ASSERT [x is fine with it], John, {John, DEP (John)}>
b. y [y ∈ {John, DEP (John)} & (y λx ASSERT [x is fine with it])].
UCLWPL 2011.01
119
Remember that the lambda operator generates utterances and not propositions as it did in the
case of focus. Therefore, the representation in (98) expresses that the speaker asserts John is fine
with it and is not in a position to assert the brother of John is fine with it. Note that the assertion
regarding John does not imply that the brother of John is not fine with it (see Neeleman and
Vermeulen (2010)) for a similar conclusion for contrastive topics), as is evident from the felicity
of the continuations of the example in (99).
Speaker A: Tell me about John‘s brother. How is he feeling lately?
Speaker B: Well, John himself is fine...
a. ...I haven‘t heard from his brother lately. He is probably fine, too.
b. ...His brother, got married recently.
(99)
The interpretation in (98) further expresses that the use of the as for reading in (90B) constrains
the members of the set of alternatives in such a way that only certain variants of the proposition
John is fine with it can be included. The position held by John in these variants can only be filled
by constituents denoting referents peripheral to John (i.e. the brother of John), as with the in
person reading.
The analysis provided here for the as for reading is in line with the hypothesis that the
intensifier should be able to be marked as contrastive topic and interact with one of the types of
alternatives we have discovered in the previous sections, namely the type that expresses, what we
called, a defining relation (on a par with the in person reading)32. All else being equal, we expect
to find another reading, which is also marked as contrastive topic but interacts with the type of
alternatives expressing a participatory relation (on a par with the alone reading). As shown in the
example below, not only do we not find a reading with such alternatives but the use of the
intensifier in a context with such alternatives is in fact impossible.
(100) John is going to Spain with Bill. #John (himself) is (himself) going to Spain (himself).
Note that the as for reading, as analysed above, has no problem with the context given in (100),
as shown below.
(101) The brother of John is going to Spain. John himself is going to Spain, too, but I don‘t
32
Note that our previous conclusion, in section 3.6, regarding the contrastively marked intensifier’s ability to
interact with an indefinite pronoun alternative referent is confirmed by the narrow down topic example below.
a)
Context: Speaker B lives with John.
Speaker A: What did some (of the people) wear at the party?
Speaker B: #Well, I didn’t go to the party after all. However, John himself wore jeans. I saw him before he
left home.
It is true that the above example can be easily accommodated by the hearer. However, the reader should note that the
judgment is according to the interpretation that John contrasts with some of the people (and of course John has no
special status against the rest of the party people). Indeed, under that interpretation the utterance in B is infelicitous.
This is confirmed by the felicity of the variation of the example above in (b), in which we do not have
intensification.
Context: Speaker B lives with John.
b) Speaker A: What did some (of the people) wear at the party?
Speaker B: Well, I didn’t go to the party after all. However, John wore jeans. I saw him before he left
home.
UCLWPL 2011.01
120
know about the rest of the family.
On further reflection, the non-existence of a reading marked as contrastive topic, which interacts
with the x with y type of alternatives should not surprise us. This is because the notion of
(contrastive) topic is intrinsically incompatible with such alternatives to begin with, as shown in
the example below. It follows that the use of the intensifier marked as contrastive topic in a
context with the comitative construction is impossible.
(102) John is going to Spain with Bill. ??As for John, he is going to France.
We think that the reason behind the incompatibility of contrastive topics, or the intensifier
marked as such, and the comitative construction type of alternative lies in the fact that
comitatives are VP modifiers. This means that the topic in the first sentence in (102) is not John
with Bill but only John. The unacceptability of the following sentence is then a result of the fact
that the topic does not change, despite the B-accent indicating that it does.
3.8. The also reading
This is one of the rarest readings found cross-linguistically. According to Gast (2009) it
represents a peculiarity of some of the European languages, including English. As can be
understood from the example in (103B), its use relates to an additive inference of the sort
someone else has also been poor, hence its frequent paraphrase with focus particles such as also
or too.
(103) Speaker A: Many people are poor, and every single one of them says it is tough.
Speaker B: Yes, I know, I have been poor myself and I can tell you it‘s not a joke.
At first sight, the use of the intensifier in (103B) does not seem to differ much from previous
uses in terms of its ability to make reference to alternatives, and more specifically to the clause in
(103A) Many people are poor. It does seem to differ from previous readings however in terms of
the fact that there is no negation33 of this alternative. Based on what was said before regarding
the function of contrast, this last feature excludes the possibility of treating this reading as an
instance of contrastive topic or contrastive focus. Assuming that the inducing of alternatives for a
given constituent indicates it being marked as a topic or focus (see discussion above), the former
feature leads us in the direction of analysing this reading along the lines of one of these two
notions. The following argument gives some direction regarding the choice among these two.
One of the most prominent features of any instance of the intensifier is its consistent
interaction with a DP. As pointed out for the uses we were concerned with thus far, the result of
this interaction leads to the interpretation of this DP itself as a constituent which is marked as a
contrastive topic or contrastive focus. Therefore, if there is some kind of DP which can be
marked only as a topic or only as a focus, and at the same time can interact with the reading we
are investigating, then we can safely conclude which of the two candidate interpretations the
intensifier carries. As Neeleman and Van de Koot (2008) point out, negative quantifiers can
33
With negation here we mean the effect of contrast. So, when contrast combines with focus it gives us
negation of an alternative proposition and when it combines with topic it gives us unwillingness to utter an
alternative.
UCLWPL 2011.01
121
function as foci but not as topics (because someone cannot say something about nothing or no
one). This is shown in examples (104) and (105) respectively. Note that in example (104), the as
for construction ensures that the constituent it introduces is marked as a topic.
(104) Speaker A: Who went to the cinema?
Speaker B: [Nobody]F went to the cinema.
(105) #As for [no boy]T, I like him.
As illustrated in (106), the also reading can in fact interact with the negative quantifier nobody34,
hence it must be the case that it is an instance of focus.
(106) Speaker A: Being poor is tough. Can any of you give me some advice on how to get
by?
Speaker B: I‘m sorry but nobody in this room has [himself]F been poor, so we cannot
advise you on that.
Being an instance of focus which is not specified with contrast, the also reading is expected not
to necessarily interact with explicit alternative propositions. But (103B) (along with every
instance of this reading, see Siemund 2000, Gast 2006) does not seem to adhere to this rule. As
pointed out above, it seems that it makes reference to Many people are poor. We think that this is
not the case, at least for the purpose of expressing its validity (truth/falsity). Instead, the
understood interaction between these two propositions is related to the additive inference of this
reading, as explained later in this subsection.
One last aspect of an intensifier with the also reading is the fact that it can only associate
with a DP whose referent is central, similarly to all the previous readings. In the case of all the
previous readings, we suggested that the intensifier can only interact with referents that have
been previously rendered as central (through an asymmetrical relation holding between the
intensified entity and some other entity). However, in this case, the context does not provide us
with identifiable alternative referents. Many people is understood as a specific indefinite DP, and
there is no way of identifying which people speaker A is referring to. Despite this, there is
nothing prohibiting the intensified referent entering into an asymmetrical relation with some
other, albeit almost (see footnote 35) unknown, referent. Since the alternative referent is not
identifiable, the also reading resorts to other ways of rendering the intensified DP as central. In
(103) for example, the intensified DP can be understood to be central through two ways (there
may be more). One, by virtue of the fact that the proposition is about the speaker himself. It is
well known (especially in the literature of logophoricity) that A is central in discourse from A‘s
point of view. Two, through the denotation of the rest of the sentence (Siemund 2000). For
example, the intensified referent may have had some experience with respect to being poor,
something which renders him/her as a central representative of this state. Therefore, the
intensified DP must still be rendered as central prior to the utterance in (103B) in order to be
felicitously intensified. However, the central status of the intensified referent in (103B) is
admittedly much less prominent when compared to a referent intensified with, for example, the
in person reading. This is confirmed by the fact that we can substitute the intensifier in (103B)
with the focus particle too or also, without any loss of felicity (note however that the native
34
This goes against Gast (2006), who suggests that this reading interacts with topical constituents.
UCLWPL 2011.01
122
speakers still interpret the presence of the intensifier as centralising the associate DP, but not
when these focus particles are present). The weak specification of centrality follows from the
fact that the also reading is not contrastively marked. Or, to put it in different terms, centrality is
understood to be weak because DEP (x), which is the expression of the asymmetrical relation (or
centrality), is not saliently evoked. This is in turn a consequence of ID (x) not being understood
contrastively. As already pointed out in section 3.6 though, this weak evoking of DEP (x) is
predicted to allow for the non-contrastively marked intensifier to interact with specific indefinite
pronouns. Indeed, this prediction is borne out by the felicity of example (103), as many people is
understood similarly to specific indefinite pronouns, as already pointed out.
Following this discussion, we propose that this reading denotes ID (x). On a par with the
rest of the readings, ID (x) interacts with the dependent function, DEP (x). The hypothesis below
summarises our analysis regarding the interpretation of this reading (note that this hypothesis
indicates only the necessary conditions for the occurrence of this reading).
(107) Hypothesis 4: The also reading of the intensifier is invariably an instance of focus and
always makes reference to dependent function(s).
In line with the representations given above for a focused constituent, a sentence such as (103B)
is analysed as follows.
(108) <λx [x has been poor], I, {I, DEP_1_ (I), DEP_2_(I), ..., DEP_n_(I)}>
The representation above expresses that the proposition I have been poor is true. It also expresses
that this proposition‘s subject referent is contained in the set of people possessing the property of
being poor. Hence, the alternative propositions of the sort an entity dependent on me has been
poor can also be true. It is also expressed that the intensified referent I is asymmetrically related
to the entities contained in the alternatives (in the alternative propositions).
It could be argued that the non-exclusion of the alternative propositions is the reason
behind the additive inference of this reading. This cannot be the whole story however. Not
negating a proposition does not necessarily mean that you are adding it. It may be the case that
this proposition is not even considered for the purpose of expressing its truth validity, as
suggested above for the also reading. Even though the non-negation of alternatives is a necessary
prerequisite for the additive inference of the also reading, the example below illustrates that it is
not sufficient. (109B) illustrates that the intensifier can interact with the constituent answering
the wh-expression of the question in (109A), hence it must be in focus (let us call this reading the
pure focus reading).
(109) Speaker A: Who did you see yesterday?
Speaker B: The King [himself]F, (walking down the road).
Since it is an instance of focus, the intensifier in (109B) does not evoke explicit alternative
propositions (in the absence of any accommodation on the part of the hearer), and consequently
cannot imply any exclusion. Note that the non-contrastive nature of the intensifier in (109B) is
further suggested by the fact that there is no need for a clarification of the sort What do you
mean? Who else did you see?, as it was the case for the contrastively marked constituents we
have seen before. This is because the notion of focus, by itself, does not force the speakers to
UCLWPL 2011.01
123
have any alternative propositions under consideration. In other words, the notion of focus does
not provide access to alternative propositions, contrary to the notion of contrast. (110B) ensures
that the intensifier with the same reading as in (109B) cannot be interpreted contrastively (for
those who find (110B) felicitous, it is because there is evoking of alternatives to the King
(contrast); this is not the reading we are aiming for) as it cannot be pre-posed felicitously in that
context (note that usually only contrastively marked elements can move; see Neeleman and Van
De Koot (2008) for a discussion).
(110) Speaker A: Who did you see yesterday?
Speaker B: #The King [himself]F I saw walking down the road yesterday.
In contrast to the also reading, the intensifier in (109B) does not relate to an additive inference,
despite being an instance of focus. Its meaning contribution is restricted in rendering the
intensified DP, the King, to be understood as central with regard to some other alternative
referents35. Therefore, it cannot be the case that the focus marking of the intensifier, by itself, is
the reason behind the additive inference observed in the also reading instances. Below we
attempt to provide an explanation of this phenomenon.
It has been observed that the also reading is extremely contextually dependent (see
Siemund 2000, Gast 2006). In order to get this reading, the proposition expressed by the sentence
containing the intensifier must be mentioned or implied in discourse36(with a different referent in
the intensified position of course). As a consequence of this, the also reading cannot be reached
when the sentence that contains it stands by itself, as shown in (111). Note that according to
native speakers, the sentence in (111) is possible when interpreting the intensifier differently (i.e.
alone, in person), something which shows that this kind of context dependence of the also
reading is special among the various instances of the intensifier.
(111) I have raised three kids [myself]F. (impossible with an also reading)
Why is it possible to interpret myself in (111) as alone/in person and not also then? These two
readings differ with respect to the fact that the former, but not the latter, is marked with contrast.
Remember that propositions with contrastively marked constituents make reference to or
saliently evoke alternative propositions which differ only in the position of the marked
constituent. In cognitive terms, marking a constituent contrastively plays to the hearer‘s ability to
accommodate the fact that there must be some alternative propositions that ought to be salient (if
35
The centrality effect is what distinguishes a sentence with a normally focused constituent and a sentence like
(109B). This difference is visible in the example below. Whilst (i) is fine in that context, (ii) is infelicitous because
John is not central in any way (in contrast to the king who is central almost by definition).
a)
Speaker A: Who did you see yesterday?
Speaker B: i) [John]F walking down the road.
ii) #John [himself]F walking down the road.
It follows from the examples above that the alternative values of an intensified constituent with the pure focus and
also reading are not completely unspecified as it is in the case of a normally focused constituent. Since the
intensified entity is defined as central, the rest of the entities in the alternative values are peripheral. In our terms,
this minor specification of the properties of the alternative values is a result of the alternative function to ID (x),
DEP (x).
36
This restriction is not specific to this instance of the intensifier. As Krifka (2007) points out, the additive
focus particles too and also also impose a similar restriction on discourse.
UCLWPL 2011.01
124
they are not, then the hearer asks for clarification). The possibility to interpret (111) as the
alone/in person reading follows from this, once we consider the fact that these readings are
interpreted as such partly because of the type of their alternatives, and crucially, the fact that they
are salient. For instance, there would be no way of interpreting (111) with the alone reading, if
we did not have in mind the fact that the alternative is a comitative construction. However, when
a constituent is not contrastively marked the hearer is not prompted to consider alternative
propositions. Since the also reading is not contrastively marked, but requires the same
predication as the predication of the proposition containing the intensifier to be present in
discourse, the hearer is not prompted to accommodate alternative propositions (which invariably
consist of the same predication), hence the impossibility to interpret (111) with the also reading.
Therefore, the notion of contrast also explains its special contextual requirements and the
impossibility of (111) (with the also reading). The fact that the also reading cannot be reached in
out of the blue contexts further substantiates the claim that it is non-contrastively focused
marked. In addition, (111) provides crucial evidence for the categorisation of the alone and in
person readings as contrastively marked instances of the intensifier. (112) illustrates that (111)
becomes felicitous with the also reading when inserted in a context containing a proposition
expressing the same predication as the predication of the proposition containing the intensifier,
raising kids.
(112) Speaker A: It‘s tough to raise kids.
Speaker B: Yes I know, I have raised three kids [myself]F and it was pretty hard.
However, we have not yet explicitly explained why the additive inference is present. We
propose that this inference is a result of the obligatory presence in discourse of the same
predication as the one contained in the intensifier‘s proposition (but with a different referent). As
we have seen in (109B), the intensifier can be an instance of focus and not convey the additive
inference. At the same time the intensifier in (109B) does not require the interlocutors to have
any other propositions under consideration (i.e. I saw an entity dependent on the king walking
down the road). On the other hand, the also reading, despite being an instance of focus,
obligatorily needs this alternative proposition under consideration because of its predication.
Assuming that a focused constituent does not need its alternative propositions to be explicitly
discourse present, the presence of an alternative proposition to the proposition containing the
also reading can only be used as a signal to the interlocutors that this alternative proposition
should be taken under consideration. Having two propositions with the same predication under
consideration, without excluding one of them, is precisely what leads to this additive inference.
Note that having the alternative proposition under consideration does not mean here that there is
access to it. It simply means that the speaker using the also reading has noted the fact that there
is already a proposition (i.e. Many people are poor) in the common ground expressing that its
predication (i.e. being poor) is applied to a referent (many people) other than the intensified one.
It is important to note that under this approach the also reading is merely a special case of the
intensifier being an instance of focus.
We end the discussion in this subsection with the type of alternatives that the pure focus
and also readings interact with. We discovered two basic types of contextual instantiations of
DEP (x) and its associate DP, namely the comitative construction expressing the participatory
relation and constructions expressing the defining relation. Unfortunately, we cannot verify the
type of alternatives the pure focus reading interacts with, as the set of alternative values that
UCLWPL 2011.01
125
includes the intensified DP is always represented by a wh-expression. On the other hand, the fact
that the also reading requires an alternative proposition to be present in discourse makes it
feasible to check what type of alternatives it interacts with. As illustrated below, the also reading
is incompatible with the comitative type of alternatives but compatible with the ones expressing
the defining relation.
(113) Speaker A: John found the way to the station with Bill.
Speaker B: #Yes, John found the way to the station [himself]F.
(114) Speaker A: The brother of John is a really good student.
Speaker B: Yes, John is a really good student [himself]F.
Should we expect another reading of the intensifier that is able to interact with the comitative
type of alternatives and simultaneously exhibit the additive inference (similarly to the also one)?
We think that the answer is no (this expectation is borne out by the infelicity of (113) in which
we keep constant the informational structural marking of the intensifier and the presupposition
that there is at least one other referent who found the way to the station, and only altered the type
of alternatives). This is due to the fact that the intensified referent constitutes a separate referent
in this kind of alternatives. As a result of this, both the proposition containing the intensifier and
its alternative proposition constitute a statement of the fact that one and the same referent
participates in the particular event described by the predication (which is the same in both
propositions). This equals to a restatement of the facts, something which is pragmatically
incoherent. This can be seen from a comparison of (113A) and (113B). Both propositions state
the fact that John found the way to the station. Note that (113B) remains infelicitous even if we
attempt to replace the intensifier with an additive particle (i.e. also, too), for the same reason.
The explanation put forward here predicts that the comitative alternative should not be able to
interact with an intensifier which requires its alternatives to be explicitly mentioned in discourse,
like in the case of the also reading, and does not deny one of them. As we will see in the next
section (3.9), this prediction is borne out by the even reading of the intensifier.
3.9. The even reading
(115a) illustrates the use of this reading. The presence of the intensifier forces the
characterization of the referent it interacts with as the least likely or expected member of a set of
individuals that could potentially replace this referent in the sentence. This scalar inference is
also exhibited by the focus particle even, hence the close paraphrase of (115a) in (115b).
(115) a. John said Einstein himself could not understand Physics sometimes.
b. John said even Einstein could not understand Physics sometimes.
Apart from the scalar inference, the even reading (and the use of the focus particle even) is also
characterized by the additive inference, similarly to the also reading. (115a) (and (115b)) is
understood as to imply that some other referents, besides Einstein, could not understand Physics
either. The difference between the use of the focus particle even and the even reading of the
intensifier lies in the way these two constituents characterize their focus. Similarly to all other
uses of the intensifier, the even reading is intuitively understood to interact with central
participants in discourse, something which is not the case in (115b). In particular, the referent of
the intensifier‘s associate DP is understood as central with respect to the topic of Physics. Based
UCLWPL 2011.01
126
on the example in (115a) it could be argued that Einstein is defined as central against the
denotation of the sentence it is contained in, namely not being able to understand Physics, and
hence the general understanding of Einstein being central with regard to Physics. Of course, this
is reminiscent of one of the ways the also reading characterizes the referent of the intensified DP
as central. Moreover, the specification of Einstein as central is equally weak as it was in the case
of the also reading. This weak specification of centrality is evident from the fact that it is
impossible to exhibit a context favoring the even reading of the intensifier but not the scalar
focus particle even (however, native speakers still perceive the even reading, but not the focus
particle even, as centralizing its associate DP). Considering these common properties of the even
and also readings (they differ only with respect to the scalar inference), it seems reasonable to
analyse the two readings likewise. For instance, we could assume that the even reading together
with its associate DP are understood as an instance of a focus marked constituent that is
understood additively because of the presence of the same predication in context. The scalar
inference is simply a result of context, and more specifically, of shared assumptions among the
discourse participants regarding the (un)likelihood of the intensified referent to form part of the
denotation of the proposition it occurs in. In (115a) for example, Einstein is judged as someone
unlikely to be unable to understand Physics, given the shared world knowledge that he is one of
the most prominent figures in the history of Physics. Attempting to replace Einstein with another
central figure, (let‘s say) Obama, would result in infelicity on the intended reading because
Obama is not a prominent figure in the field of Physics. If the scalar effect is a pragmatic
phenomenon, then we should be able to eliminate it just by inserting (115a) in a context where
Einstein is not an unexpected or unlikely referent, given the rest of the proposition and the
alternative referents contained in the set. (116) illustrates that this is possible.
(116) Speaker A: As far as I know, Isaac Newton and Stephen Hawking found it tough to
understand Physics at first.
Speaker B: Yes, I know, Einstein himself found it tough to understand it sometimes.
(116) shows that by simply manipulating the set of alternative referents that are contextually
present and compared to Einstein we are able to eliminate the scalar effect. World knowledge
instructs us that Newton and Hawking are also prominent figures of Physics. The fact that they
have also been in the situation of finding it tough to understand Physics removes the
(un)likelihood of something similar happening in the case of Einstein. Note that, when the scalar
effect is eliminated, as in (116), the most accessible reading that we are left with is the also one.
This is certainly expected under this analysis, since it is assumed that the two readings are
essentially the same and only differentiated in terms of the scalar inference, whose presence (or
absence) is dependent on the set of alternative referents provided by the context.
However, a closer comparison of the even and also readings suggests that the scalar
inference cannot be the only factor distinguishing them. As pointed out in the discussion
surrounding (111), repeated below as (117), the also reading is inaccessible when the proposition
containing it does not occur in a context where there is another sentence containing the same
predication (for (117) this predication must be along the lines x has raised y number of kids).
(117) #I have raised three kids [myself]F.
UCLWPL 2011.01
127
As pointed out, this resulted from the fact that the also reading is a non-contrastively focused
constituent that cannot evoke salient alternative propositions. Since it cannot evoke alternative
propositions, its requirement to occur in a context in which the same predication as the one it
occurs in is present remains unsatisfied (remember that the alternatives to a sentence containing
an intensifier differ only with respect to the position occupied by the associate DP and the
absence of the intensifier; the rest of the predication remains the same). Contrary to these
contextual restrictions of the also reading, which render (117) infelicitous, the even reading is
easily accessible in (115a). Hence, it cannot be the case that it is the same as the also reading
with the scalar inference added because of pragmatics. Despite the fact that the predication
contained in the sentence in (115a), namely could not understand Physics, is not present
contextually, we can still access the even reading and the additive inference that comes with it.
Given that a) this additive inference is a result of considering the fact that the same predication is
applied to some other referent, and b) the even reading is able to exhibit this additive inference in
a context deprived of this predication, the even reading must be able to realize the consideration
of the same predication (applied to some other referent) through other means. We propose that
the even reading‘s ability to do this is a result of it and its associate DP being interpreted
contrastively. The rationale behind this assumption lies in the fact that contrast provides access
to the induced alternatives (see section 3.2). Since these alternatives consist of the same
predication as the sentence containing the intensifier, we can explain why the even reading does
not require this predication to be explicitly expressed in context in order to be accessed. Note
that this reading and its associate DP must be an instance of contrastive focus, and not
contrastive topic, because the associate DP can be a negative quantifier as shown below.
(118) Context: Some of the members of the Olympics football team have been caught on
steroids in a pre-Olympic steroid test. As a result, the whole team now has a bad name.
Speaker A is the assistant manager and speaker B is the manager of the team.
Speaker A: The team is doing really well. I think we will do well in the Olympics.
Speaker B:Yes, I know, but no loss would be a bad result ITSELF . I bet everyone will
start saying that we won all our matches because we are on steroids. (paraphrase: ...
even no loss would be a bad result...).
Given these considerations, we propose the following.
(119) Hypothesis 5: The even reading of the intensifier is invariably an instance of contrastive
focus and always makes reference to dependent function(s).
The reference to the dependent function, DEP (x), will deliver the intuitively correct alternative
referents (consisting of a central entity and peripheral ones) contrasting with the referent of the
intensified DP, similarly to the rest of the readings. Further evidence for the contrastive character
of the even reading and its associate DP comes from their distributional behaviour. Before (see
sections 3.3 and 3.8), we touched upon the fact that constituents marked with contrast exhibit
certain grammatical effects. In particular, as we have seen in example (31), repeated below as
(120), the contrastively focused constituent The Selfish Gene can be moved to the front of the
clause.
(120) Speaker A: John read The Extended Phenotype.
UCLWPL 2011.01
128
Speaker B: No, you are wrong, THE SELFISH GENE he read.
Wh-questions do not normally license focus movement in the answer, as the example in (52),
repeated below as (121), shows. This is because The Selfish Gene is (focused but) not
contrastively marked.
(121) Speaker A: What did John read?
Speaker B: #The Selfish Gene he read.
The hypothesis in (119) predicts that the even reading along with its associate DP should be able
to move, in parallel fashion to the rest of contrastively marked constituents. This prediction is
borne out as illustrated in (123), in which the constituent Einstein himself can be pre-posed
felicitously. Example (122) shows that this is not a peculiarity of the even reading of the
intensifier, but in fact every constituent that denotes an unlikely referent with respect to the
denotation of the sentence can be pre-posed felicitously.
(122) Even EINSTEIN John said could not understand Physics sometimes.
(123) Einstein HIMSELF John said could not understand Physics sometimes
The hypothesis in (119) has proved sufficient to explain why the even reading and its associate
DP can move, as well as certain interpretative aspects of this use. However, there is an obvious
incompatibility of the interpretation of the notion of contrastive focus and the interpretation of
the even reading (or the focus particle even). Until now, we have been assuming that the
proposition that contains a contrastively focused constituent evokes alternative propositions and
simultaneously negates (at least) one alternative proposition. On the other hand, we described the
even reading‘s interpretation as evoking alternative propositions and ranking them in terms of
likelihood or expectancy, out of which the least likely one is selected, hence the scalar
proposition. Crucially, there is no negation of any alternative propositions. After all, if this were
the case, the additive inference that characterizes this reading would be impossible. In order to
overcome this problem, which lies in the core of our general proposal, we adopt a different
conceptualization of the notion of contrast than that described by Neeleman and Vermeulen
(2010) (see section 3.2). In particular, following Frey (2010), we perceive the function of the
notion of contrast to be the ordering of the set of alternatives according to various criteria
including truth value, likelihood, and expectancy. In practice, this does not change much. What
we perceived until now as being the function of a contrastively marked constituent (denial of
alternatives) is now a subcase of the largest family of phenomena involving this notion, whose
effect is the ordering of the alternatives in the relevant set according to some criterion. For
instance, in corrective contexts or in the cases of the in person, alone and as for readings, the
notion of contrast ranks the propositions/utterances contained in the set according to the criterion
of truth value. In (120) for example, speaker B expresses that the proposition he/she expresses is
ranked as true and the alternative proposition expressed by speaker A is ranked as false in a scale
of truth. This interpretation is essentially the same as the one expressed by Neeleman and
Vermeulen (2010), hence there is no problem with the analysis of all the previous instances of
the intensifier involving the notion of contrast.
UCLWPL 2011.01
129
Once we adopt this view of the notion of contrast, the scalar inference of the even reading
falls out naturally. Also, the additive inference expressed by this reading is not incompatible with
the notion of contrast anymore. The intensifier and its associate DP are indeed interpreted as a
contrastive focus, however the presence of contrast, instead of ordering the propositions
contained in the given set according to the criterion of truth value, it orders them in terms of the
criterion of likelihood or expectancy. Hence, the interpretative effect of the even reading, namely
to render the referent of its associate DP as the least likely or expected candidate with respect to
the rest of the denotation of the proposition, is an effect of contrast ordering the set of referents
in a scale of likelihood or expectancy. Below, we provide the formalization of the interpretation
of the even reading in (115a) (we leave out parts of the sentence irrelevant to the point)3738.
(124) a. <x[x could not understand Physics], Einstein, {DEP_1_(Einstein),
DEP_2_(Einstein), ..., DEP_n_(Einstein)}>
b. x [x  {DEP_1_(Einstein), DEP_2_(Einstein), ..., DEP_n_(Einstein} 
[Einstein could not understand Physics] <c [x could not understand Physics] ].
(124) expresses to what extent the set of contextually relevant individuals (i.e. Einstein, someone
studying Physics, etc) is contained in the set of individuals that could not understand Physics. It
is asserted that one member of this set, namely Einstein, is also a member of the set of
individuals that could not understand Physics. In addition, it is asserted that, given the common
ground c, Einstein is less likely not to understand Physics than any other member of the
contextually given set. The presence of alternatives and the first assertion are the result of the
semantics of focus, whereas the second assertion corresponds to the semantics of contrast.
Contrary to any other contrastively focused constituent with this particular semantics, however,
the even reading of the intensifier in (115a) constrains the members of the set of alternatives in
such a way that only certain variants of the asserted proposition can be included. The position
37
<c indicates that the proposition on the left of it is less likely to be true than the one on the right of it, given
the common ground c.
38
It is true that the semantics of the type of contrast that ranks the alternatives in terms of likelihood differ
considerably from the semantics of the type of contrast that ranks the alternatives in terms of truth values. Hence, it
could be argued that subsuming both instances of semantics under the notion of contrast is something which is
conceptually satisfactory but nevertheless does not follow from the level of semantics. Formally, it is hard to
maintain the conclusion that the two instances of semantics are representations of the same notion (contrast). Setting
aside the arguments (i.e. movement) tying together the notions represented by the two different instances of
semantics under the label of contrast, as spelled out above, we would like to offer an informal rationale for this
discrepancy at the level of semantics. The two instances of semantics differ in two respects. In the case of the truth
value ranking there is negation of at least one alternative, whereas in the case of the likelihood ranking there is the
assertion that all the alternatives are more likely than the one asserted. Hence, there is difference in the type of
assertion (negation vs more likely) and the reference to alternatives (all vs at least one). We think that the latter
property is a direct result of the former. In order to be able to make an assertion regarding the least likelihood of a
given alternative, there must be reference to the more likelihood of all the alternative propositions in context. It
cannot be asserted that an alternative is the least (notice the superlative) likely just by making reference to the more
likelihood of some alternatives, hence the use of the universal quantifier. On the other hand, negating alternatives is
not as demanding in terms of making reference to alternatives. The requirement to negate is still satisfied by making
reference to one alternative, hence the use of the existential quantifier. Of course both instances of contrast involve
opposing of alternatives, and that is precisely the function of contrast as we have been arguing up to now. At the
level of semantics this translates into the fact that we get a second assertion (the (b) part in every semantics that
involve contrast).
UCLWPL 2011.01
130
held by Einstein in these variants can only be filled by constituents denoting referents peripheral
to Einstein (i.e. someone studying Physics), and nothing else.
Following the new view on the notion of contrast we adopted above, the question arises as
to which particular instantiation of the semantics of contrast is relevant in a given instance of the
intensifier. Or, to put it differently, what makes the hearer decide according to which criterion
(i.e. truthfulness, likelihood) the speaker contrastively marks a particular constituent? The
answer to this question will be the same as the one given in section 2, regarding the choice of the
interpretation of the intensifier. The main factor distinguishing which interpretation of the
intensifier is selected is context. According to our theory, both the in person and even readings
are contrastively focused, but with different semantics, hence a sentence like (115a) should be
able to take either interpretation of the two (in fact, it can take more than these two readings,
again depending on the context; see for instance example (116)). The reason that the even
reading is preferred here is the predication it occurs in, could not understand Physics, which
forms part of the context of course (hence the contribution of pragmatics, as outlined at the
beginning of this subsection, is indeed significant). As shown below, (115a) can be interpreted as
the in person reading when inserted in different context (corrective context) that overrides the
effect of the predication.
(125) Speaker A: John said Einstein‘s son could not understand Physics sometimes.
Speaker B: No, John said Einstein HIMSELF could not understand Physics sometimes.
The negation used by speaker B, characteristic of corrective context, is an indicator towards the
hearer to interpret the contrastively focused constituent according to the criterion of truthfulness
(see also Frey 2010 for a similar point), and hence the interpretation of the intensifier as in
person (and the respective semantics).
Being marked as contrastive focus, we would expect the even reading to exhibit the same
requirements (as the rest of the readings marked as such) regarding the alternative values it
interacts with, namely to not be able to interact with indefinite pronouns found in certain
configurations (see for instance example (88) illustrating this for the in person reading).
However, as illustrated below, this is not the case as it can interact with many people, which is
arguably interpreted as a specific indefinite pronoun, similarly to the also reading (a noncontrastively marked reading).
(126) Many people could not understand Physics. Einstein HIMSELF could not understand
Physics sometimes.
What differentiates the two instances of contrast is the criteria according to which the
alternatives are ranked. Whereas the in person, alone and as for readings rank their alternatives
in terms of truth value, the even reading ranks them in terms of likelihood or expectancy. This is
evident from the semantics of each instance of contrast. According to the semantics of the former
instance, there is negation of an alternative proposition/utterance. In the semantics of the latter
instance, it is stated that the proposition expressed is the least expected from the relevant set.
Notice that no direct reference to, or assertion regarding, alternative propositions is made in this
latter instance of contrast. The part of the semantics which expresses unlikelihood is primarily
concerned with the proposition expressed by the speaker, and the fact that, given the context, this
proposition is an unlikely candidate. This semantics matches our intuition regarding the function
UCLWPL 2011.01
131
of this instance of contrast. When uttering (126), we have the general understanding that Einstein
is an unlikely individual to not understand Physics. The reason for this is because Einstein is a
Physicist with many achievements. What will come to mind is the fact that Einstein is a great
Physicist, and because of this he is unlikely to have not been able to understand Physics. Notice
that no reference to other individuals is made but only to the reason behind this unlikelihood. Of
course there is a sense of comparison between Einstein and other individuals, which results from
the common assumption made by speakers that for someone to not understand Physics, he/she
should have at least less abilities in Physics compared to Einstein. Since other individuals could
not understand Physics, their level of understanding Physics must have been at least lower than
Einstein. On the other hand, in the case of contrast specified in terms of the scale of truth, there is
a clear assertion regarding an alternative proposition/utterance. As pointed out before (see
section 3.3), in order to be able to do this, clear access to alternatives is required.
In a nutshell, the two types of contrast behave differently in terms of their access to
alternative propositions. We can say that the negation type of contrast gives full access to
alternatives, because there must be negation of one of them39, whereas the likelihood type of
contrast gives limited access to alternatives, because there is only some kind of indirect
comparison between them (and for this comparison to happen it is sufficient to know the
individual that is intensified and the criterion of this comparison). It follows that the even reading
and its associate DP are not predicted to impose similar (to the rest of the contrastively marked
readings) restrictions regarding the identifiability of the interacting alternative referents. This is
because contrast defined in terms of likelihood does not provide clear access to alternatives, and
hence DEP (x) is not saliently evoked. This also explains the otherwise unexpected weak
centrality effect imposed by this reading on its associate DP.
In section 3.8 we were unable to find a reading of the intensifier exhibiting an additive
inference and at the same time interacting with the comitative type of alternatives. As explained
(see section 3.8), this is due to the fact that the entity that is subsequently intensified constitutes a
separate referent in the alternative. As a consequence, we pointed out, an intensifier which does
not deny the salient alternative containing the comitative construction cannot interact with the
central entity in this construction because it would constitute a (partial) repetition of the
alternative proposition already expressed. What is expected then is that there should not exist a
reading with the scalar inference interacting with the comitative type, because the scalar
inference does not deny the evoked alternatives. In fact scalarity encompasses the addition of
alternatives. Needless to say that we also expect the even reading of the intensifier and its
associate DP to contrast with the type of alternatives that express the defining relation (similarly
to all the previous readings except the alone one), and not the comitative type that expresses a
the participatory relation. Indeed, as shown below in (127) and (128), an intensifier characterized
with the scalar inference cannot interact with type of alternatives expressing the participatory
relation but only with the ones expressing the defining relation. This interaction delivers the even
reading.
(127) Context: John is not good with finding his way around.
Speaker A: John found the way to the station with Bill.
Speaker B: #So what? Finding the station is not hard nowadays. John (HIMSELF ) found
39
This is in accordance to what has already been pointed in section 3.3, namely that the saliency of alternatives
(propositions or utterances) is a by-product of their negation.
UCLWPL 2011.01
132
the way to the station (HIMSELF ).
(128) Context: John is not good with finding his way around.
Speaker A: John‘s brother found the way to the station.
Speaker B: So what? Finding the station is not hard nowadays. John (HIMSELF ) found
the way to the station (HIMSELF ).
Note that (127) remains infelicitous even if we attempt to replace the intensifier with the scalar
focus particle even, for the same reason that the intensifier is infelicitous.
According to the new view of the notion of contrast adopted in this section, whenever the
intensifier is marked as contrastive focus, the realisation of contrast can have two different
semantics. These semantics are defined in terms of the scale of truth or the scale of likelihood.
The in person reading is an instantiation of the former scale and the even reading an instantiation
of the latter, both being a contrastively focused constituent though. Based on this approach, one
would expect that there should exist at least one reading marked as contrastive topic realising the
semantics of the scale of truth and another one realising the semantics of the scale of likelihood.
The as for reading is an instantiation of the former scale. To our knowledge, a reading of the
intensifier which expresses some sort of scaling in terms of likelihood and is simultaneously
marked as contrastive topic does not exist. The reason for this is the general incompatibility of
the notion of topic and the ranking of alternatives in terms of a scale or likelihood/expectancy
(no one talks about emphatic topic or scalar topic, but only about emphatic focus and scalar
focus). We remain agnostic as to why this is the case.
3.10. The aboutness reading
As pointed out before, owing to the existence of the notions of contrast, topic and focus a
constituent in English can be marked with the following four ways: a) contrastive focus, b)
contrastive topic, c) focus, d) topic. Until now, we have seen the intensifier being an instance of
the first three notions. The theory proposed here predicts the existence of another reading of the
intensifier, namely that which is an instance of topic. The example given below confirms this
prediction.
(129)
Context: It’s the royal wedding and the Prime Minister along with other people are
walking past in a procession. A BBC commentator is describing the various outfits
worn by the PM and the rest of the people.
a. The PM [himself]T is wearing a blue suit and a yellow tie.
b. As for the PM [himself]T, he is wearing a blue suit and a yellow tie.
c. [The PM]T is wearing a blue suit and a yellow tie.
(129b) is added in order to ensure that the constituent The PM himself in (129a) can be
introduced by the construction as for. As pointed out before, the constituents introduced by this
construction are always topics. (129c) is added for comparison purposes. All three possible
utterances of the commentator converge to the point that they introduce the PM as the topic of
discourse and provide new information about him/her.
Since it is an instance of topic, the intensifier in (129a-b) renders the sentence it is
contained in to be associated with a set of alternative utterances, which differ only in the value of
UCLWPL 2011.01
133
the position occupied by the DP the PM himself. However, since this reading is not contrastive,
there is no access to these alternatives as is evident from the fact that when (129a-b) are uttered,
alternative utterances about the outfits of other guests (i.e. the PM‘s secretary is wearing a
yellow dress) do not necessarily come to mind40. In contrast to a constituent marked as
contrastive topic, the hearer does not get the impression that the speaker is unwilling (or unable)
to make alternative assertions about someone else. This is confirmed by the fact that there is not
a trigger of a request for clarification on behalf of the hearer (i.e. Who else is wearing what?).
The difference between a normal topical construction like (129c) and a topical intensified
construction like (129a-b) lies in the fact that the latter characterises the intensified referent, the
PM, as central, against some alternative referents41. For the case of (129a-b), this centrality effect
is achieved through world knowledge regarding the high political status entailed by the position
of the PM.
The peripheral referents do not have to be identifiable though (i.e. the PM‘s wife, the PM‘s
secretary etc). The non-identifiability of the alternative referents follows from the fact that the
intensifier is not marked with contrast, as explained in previous sections, which in turn makes it
felicitous in a context like (129), in which there are no explicit alternatives42. In accordance with
the suggestion made for the case of the focus readings, the nature of alternative referents this
reading interacts with is a result of the function it denotes, ID (x), interacting with the dependent
function DEP (x). Therefore, we assume the following. In order to be able to distinguish this
reading from the rest, let us call it the aboutness reading.
(130) Hypothesis 6: The aboutness reading of the intensifier is invariably an instance of topic
and always makes reference to dependent function(s).
Since the aboutness reading is an instance of topic, a sentence like (129a) is analysed as in (131).
Note that the representation below is in accordance with the intuition (as it was for the case of a
40
This is a similar case to the focus readings we have seen above. Even though it is assumed that there is
inducing of alternative propositions, these propositions are not accessed.
41
This is why the commentator can utter something like (a) below, and not (b) (same context as in (129)).
Mary cannot be intensified in (b) because it is not central to any other alternative referents in that context.
a) [Mary]T is wearing a blue dress.
b) #Mary [herself]T is wearing a blue dress.
Note that the centrality effect in this reading is equally weak as in the case of the pure focus, also and even readings.
As the reader may have noticed however, in the case of the pure focus and aboutness readings we have been able to
give examples showing that they are infelicitous because of this centrality requirement. We have not been able to do
something similar for the also and even readings. Despite this we still think that the specification of centrality is in
all these readings equally weak. The reason for not being able to give infelicitous examples in the cases of the also
and even readings is because there is always an alternative proposition induced, or present, in discourse with the
same predication, and importantly a referent being part of it. This referent is immediately accommodated as
peripheral to the intensified based on the rest of the denotation of the proposition (the predication). Hence, it is
impossible to provide an infelicitous example (however native speakers are positive about the existence of centrality
in these readings). On the other hand, in the cases of the aboutness and pure focus readings, we can verify the
existence of centrality because there are no alternative referents present in discourse. Therefore, the referent resorts
into other (than the predication it occurs in) ways to be central (i.e. through social/political status). This allows us to
give infelicitous examples with referents that are not generally considered as central.
42
Similarly to the case of the pure focus reading of the intensifier, the fact that the alternative referents are not
explicit in discourse makes it impossible to check what type of alternatives (i.e. comitative) the aboutness reading of
the intensifier and its associate DP interact with.
UCLWPL 2011.01
134
normal topical constituent) that speaker B performs the following speech acts when uttering
(129a): a) Consider the PM out of a set of possible topics (i.e. an entity dependent on the PM),
b) I assert that the PM is wearing a blue suit and a yellow tie.
(131) <λx ASSERT [x is wearing a blue suit and a yellow tie], the PM, {the PM, DEP_1_
(PM), DEP_2_ (PM), ..., DEP_n_ (PM)}>
We stress the point that the above representation differs from a topical constituent with regard to
the alternative utterances induced. When the intensifier is present, the constituents replacing the
intensified DP in the alternative utterances are peripheral to the intensified referent, precisely due
to the application of the dependent function.
3.11. The incompatibility of the intensifier and true indefinites
In section 2 we pointed out that all the readings of the intensifier cannot be associated with true
indefinite DPs. We concluded that the intensifier can in fact interact with indefinite DPs, but
these DPs are obligatorily interpreted as generic or specific as illustrated with examples (19b),
(20), (23), (24) for the in person, alone, also/even, and as for readings respectively. (132) – (135)
below illustrate that this conclusion is also valid for the rest of the readings we discussed. In
(132), we have an instance of the delegative reading. The only way of felicitously interpreting
the sentence is when there is a specific woman participating in the action described by the
predicate. In (133) we also have a specific reading of the indefinite DP interacting with the
without x’s help reading. In (134 the intensifier takes the aboutness reading and is interpreted as
generic. In (135) the intensifier takes the pure focus reading and is interpreted as specific43.
(132) A woman dyed her hair HERSELF.
(133) A general has commanded the whole army HIMSELF.
(134) Context: The lecturer begins his lecture on democracy with the sentence below
A politician [himself]T is the main representative of democracy.
(135) Context: Politicians are considered very important figures in the society that the
following conversation takes place.
Speaker A: Who did you see yesterday?
Speaker B: A politician [himself]F walking down the road.
As we have seen, the (only) common property of all the readings of the intensifier is the central
status of its associate DP. Therefore, we expect that this property should be held responsible for
the inability of the intensifier to interact with true indefinites. As pointed out before, the
centrality status of the intensified DP is nothing else than the asymmetric relation holding
between the entity denoted by this DP and another entity, in the alternatives. Having this in
mind, we can now understand why definite DPs and specific and generic indefinite DPs can
interact with the intensifier. Definite and specific indefinite DPs denote a particular individual;
43
Remember that there is no reading of the intensifier that interacts with a coordination alternative (the reading
that we call only). Hence, its inability to render an indefinite DP generic or specific, as shown in footnote 14.
UCLWPL 2011.01
135
therefore there should be no problem with them entering into a relation with some other
individuals. Generic DPs on the other hand denote a group/set of individuals in the world that
have some properties in common. It is conceivable for a group to enter into a relation with some
other group or referent (in (135) for instance, the group of politicians are defined as central with
respect (thus related) to the rest of the citizens in the context of democracy) hence the possibility
of generic indefinite DPs to be intensified.
True indefinite DPs, however, denote a bundle of properties and not a particular individual
in the world. (136) for example, expresses that the subject referent wants to buy an entity that
consists of certain properties (i.e. it has four wheels), a car. The phrase a car can be seen as a
way of grouping these properties.
(136) I want to buy a car.
Since one of the intensifier‘s basic functions is to interact with central referents, and since
centrality is an asymmetric relation holding between the intensified entity and another entity
(within the alternative values), a true indefinite DP cannot be intensified for the reason that its
denotation cannot be related to something else. This is because the very concept of relation
requires, by definition, at least two individuals to ‗connect‘ them together. This way of thinking
also explains why the intensifier, in contrast to other constituents serving similar functions such
as the focus particles (i.e. even, only), cannot interact with predicates (i.e. verbs), once we adopt
the well-accepted assumption that predicates denote a bundle of properties, similarly to true
indefinite DPs.
4. Conclusion
The analysis developed in this paper does not only account for the well-documented readings of
the intensifier (i.e. in person, alone, also) but goes one step further in predicting the existence of
more readings, which have, to our knowledge, remained undiscovered (i.e. aboutness).
Following Eckardt‘s (2001) proposal about the denotation of some of the readings of the
intensifier, we suggested that every instance of the intensifier has the same denotation (ID (x)).
The interpretational differences between the different readings essentially reduce to one factor,
namely the nature of the intensifier‘s marking (with the notions comprising a specific component
of information structure; see table in (27). We think that this theory is superior to previous ones
in terms of accounting for the interpretation of the intensifier for the following reasons;
(137) a. It is superior from an explanatory point of view in the sense that it assumes only
one lexical entry for the intensifier and derives its different readings from an
independently well-studied phenomenon.
b. It is flexible enough to account for the existing data in the literature (descriptive
adequacy), and furthermore predicts, correctly, the existence of more instances of
the intensifier (predictive power).
c. It is restrictive enough to limit, through the employment of the alternative function
DEP (x), the kind of alternatives the intensifier can interact with, and hence the
possible readings we can find.
d. It exposes how other factors (i.e. syntactic position, pragmatics) may come into
play for the realisation of a particular reading.
UCLWPL 2011.01
136
The table below summarizes our findings with regard to the different readings of the
intensifier44.
Reading
Denotes
Interacts with Marked as
In person, delegative, even, ID (x)
alone, without x‘s help
DEP (x)
CF
As for
ID (x)
DEP (x)
CT
Also, pure Focus
ID (x)
DEP (x)
F
Aboutness
ID (x)
DEP (x)
T
References
Baker, C. (1995). Contrast, discourse prominence, and intensification, with special reference to locally free
reflexives in British English. Language , 73.1, 63 - 101.
Bergeton, U. (2004). The Independence of Binding and Intensification (Ph.D. thesis). University of Southern
California.
Browning, M. A. 1993. Adverbial Reflexives. NELS 23.1: 83-94.
Büring, D. (2001). Let‘s phrase it! – focus, word order, and prosodic phrasing in German double object
constructions. In G. Müller, & W. Sternefeld (Eds.), Competition in Syntax, no. 49 in Studies in Generative
Grammar (pp. 101 – 137). Berlin & New York: de Gruyter.
Chafe, W. L. (1976). Givenness, contrastiveness, definiteness, subjects, topics and points of view. In C. N. Li (Ed.),
Subject and topic (pp. 25 - 55). New York: Academic Press.
Cohen, D. (1999). Towards a unified account of intensive reflexives. Journal of Pragmatics , 31, 1041 - 1052.
Constantinou, H. (2011). The Syntax of IDensification. Ms. UCL
Dirven, R. (1973). Emphatic and reflexive in English and Dutch. Leuvense Bijdragen , 63, 288 – 299.
Eckardt, R. (2001). Reanalysing selbst. Natural Language Semantics , 9(4), 371 -412.
Edmondson, J., & Plank, F. (1978). Great expectations: an intensive self analysis. Linguistics and Philosophy , 2,
373 – 413.
Féry, C. (2010). Prosody and information structure of the German particles selbst, wieder and auch. In T.
Borowskyi, K. Shigeto, S. Takahito, & S. Mariko (Eds.), Prosody Matters: Essays in Honor of Elisabeth O.
Selkirk. London: Equinox Press.
Fraurud, K. (1990). Definiteness and the processing of noun phrases in natural discourse. Journal of Semantics , 7,
395–433.
Frey, W. (2010). A-bar-Movement and conventional implicatures: About the grammatical encoding of emphasis in
German. Lingua , 1416 - 1435.
Gast, V. (2009). A contribution to ‘two-dimensional’ language description: the Typological Database of Intensifiers
and Reflexives. Ms. Freie Universität Berlin.
Gast, V. (2006). The Grammar of Identity. Intensifiers and Reflexives in Germanic languages. New York:
Routledge.
Hankamer, J. (1973). Unacceptable ambiguity. Linguistic Inquiry 4: 17–68.
Hawkins, J. A. (1978). Definiteness and Indefiniteness: A Study in Reference and Grammaticality Prediction.
London: Croom Helm.
Hawkins, R. (1981). Towards an account of the possessive constructions: NPs and the N of NP. Journal of
Linguistics , 17, 247—269.
Hole, D. (2002). Agentive selbst in German. In G. Katz, S. Reinhard, & P. Reuter, Sinn und Bedeutung VI –
Proceedings of the 6th Annual (pp. 133 - 150). Osnabrück: Institute of Cognitive Science.
Jackendoff, R. (1972). Semantic Interpretation in Generative Grammar. Cambridge: Cambridge University Press.
Kiss, K. (1998). Identificational focus versus information focus. Language , 74, 245 - 273.
44
CF= contrastive focus, CT= contrastive topic, F=focus, T=topic
UCLWPL 2011.01
137
König, E. (2001). Intensifiers and reflexives. In M. Haspelmath, E. König, W. Oesterreicher, & W. Raible (Eds.),
Language Typology and Language Universals – An International Handbook of Contemporary Research (pp.
747 - 760). Berlin & New York: Mouton.
König, E. (1991). The Meaning of Focus Particles: A Comparative Perspective. London: Routledge.
König, E., & Siemund, P. (1999). ‗Intensifiers and reflexives: A typological perspective. In Z. Frajzyngier, & T.
Curl (Eds.), Reflexives. Forms and Functions (pp. 41-74). Amsterdam: Benjamins.
König, E., & Siemund, P. (2011). Intensifiers and Reflexive Pronouns. The World Atlas of Languages online:
http://wals.info/ (date accessed: 27/3/2011).
Koptjevskaja-Tamm, M. (2004). ―Maria‘s ring of gold‖: adnominal possession and non-anchoring relations in the
European languages. In J. Y. Kim, Y. Lander, & B. H. Partee (Eds.), Possessives and Beyond: Semantics and
Syntax (pp. 155 - 181). Amherst, MA: GLSA Publications.
Koptjevskaja-Tamm, M. (2000). Romani genitives in cross-linguistic perspective. In V. Elsik, & Y. Matras (Eds.),
Grammatical Relations in Romani: The Noun Phrase (pp. 123 – 149). John Benjamins.
Krifka, M. (2007). Basic Notions of Information Structure. In C. Fery, & M. Krifka (Eds.), Interdisciplinary Studies
of Information Structure 6. Potsdam.
Lambrecht, K. (1994). Information Structure and Sentence Form. Cambridge: Cambridge University Press.
Langacker, R. W. (1995). Possession and possessive constructions. In J. R. Taylor, & R. E. MacLayry (Eds.),
Language and the Cognitive Construal of the World, Trends in Linguistics, Studies and Monographs (pp. 51 –
79). Berlin: Mouton de Gruyter.
Langacker, R. W. (1993). Reference-point constructions. Cognitive Linguistics , 4, 1 - 38.
Larson, R. (1988). On the Double-Object Construction. Linguistic Inquiry , 19, 335 - 391.
Lehmann, C., & Shin, Y. M. (2005). The functional domain of concomitance. A typological study of instrumental
and comitative relations. In C. Lehmann (Ed.), Typological Studies in Participation. Language Typology and
Universals (pp. 9 - 104). Berlin.
Molnar, V. (2002). Contrast – from a contrastive perspective. In H. Hallelgard, S. Johansson, B. Behrens, & C.
Fabricius-Hansen (Eds.), Information Structure in a Cross-Linguistic Perspective (pp. 147 - 161).
Amsterdam/New York: Rodopi.
Moravcsik, E. (1972). Some crosslinguistic generalizations about intensifier constructions. CLS, 8, pp. 272 – 277.
Neeleman, A., & Van de Koot, H. (2008). Dutch scrambling and the nature of discourse templates. Journal of
Comparative Germanic Linguistics , 11 (2), 137 – 189.
Neeleman, A & Vermeulen, R (2010). Introduction. In Ad Neeleman & Reiko Vermeulen
(eds.), A flexible theory of topic and focus movement. Ms. UCL.
Nikolaeva, I., & Spencer, A. (2010). The Possession-Modification Scale: A universal of nominal morphosyntax. Ms.
Ning Zhang, N. (2007). The syntax of English comitative constructions. Folia Linguistica, 41/1–2, 135 – 169.
Primus, M. (1992). Selbst – variants of a scalar adverb in German. In J. Jacobs (Ed.), Informationsstruktur und
Grammatik (pp. 54–88). Opladen: Westdeutscher Verlag.
Reinhart, T. (1981). Pragmatics and Linguistics: An analysis of sentence topics. Philosophica , 27, 53-94.
Rizzi, L. (1997). The fine structure of the left periphery. In L. Haegeman (Ed.), Elements of Grammar. Dordrecht:
Kluwer.
Rooth, M. (1992). A theory of focus interpretation. Natural Language Semantics , 1, 75 – 116.
Rooth, M. (1985). Association with Focus (Ph.D. thesis). University of Massachusetts at Amherst .
Sayers, D. L. (1986). Murder Must Advertise. New York: Harper and Row.
Selkirk, E. (1984). Phonology and Syntax: The Relation between Sound and Structure. Cambridge: MIT Press.
Selkirk, E. (1996). Sentence prosody: Intonation, stress and phrasing. In J. A. Goldsmith (Ed.), The Handbook of
Phonological Theory. Oxford: Blackwell.
Siemund, P. (2000). Intensifiers in English and German. A comparison. New York: Routledge.
Tavano, E. M. (2006). A bound-variable analysis of the adverbal emphatic reflexive, or How I wrote this paper
myself. Ms. University of Southern California .
Vallduví, E. (1992). The Informational Component. New York: Garland.
Vallduví, E., & Vilkuna, M. (1998). On rheme and konstrast. Syntax and Semantics , 29, 79 - 108.
Vermeulen, R. (2010). The syntax of topic, contrast and contrastive topic in Japanese and Korean. Short paper
prepared for the presentation at the On Linguistics Interfaces II (OnLI II) conference at the University of
Ulster.
Download