Temporal homogeneity and the Italian Perfect*
Fabrizio Arosio
1. Introduction
In the literature about Italian, we find little agreement on the very basic
meanings conveyed by Perfect sentences and if we look at more theoretical
analyses, even this modicum of agreement disappears. One of the reasons
for this confusing picture comes from the (vain) search for a nonambiguous semantics of the “Perfect” which accounts uniformly for the
different meanings conveyed by Perfect constructions (Bertinetto 1986,
Giorgi and Pianesi 1997). In this paper I will explore a different perspective
by looking at the Italian Perfect as authentically ambiguous (von Stechow
2002) between different meanings.
The different meanings associated with the Italian tense forms have
been analyzed either compositionally, in terms of covert aspectual operations localizing the described eventuality within or including a relevant interval introduced by the semantic tense (von Stechow 2002), or, more generally, in terms of different structures of quantification over events and/or
times inherently associated with them (Giorgi and Pianesi 1997, 2000;
Bonomi 1997, 1999). I will show that if we consider the distribution of durative adverbials this is not enough. I will propose that Italian tenses impose some conditions concerning the temporal homogeneity (Bennett and
Partee 1978) of their complements and, in particular, that Perfect constructions should fulfill these conditions. In order to implement this idea, I will
decompose the tense projection into a temporal variable, localizing the described eventuality in the flow of time, and an homogeneity condition on
the interpretation of the tense complement. The idea is that tense looks at
its complement and licenses it if it satisfies a condition of temporal homogeneity: this is to say that tense itself has some influence in the aspectual
interpretation of a sentence.
2. Italian facts
In Italian, durative per- and da-adverbials are found in complementary distribution in state sentences1:
38
Fabrizio Arosio
per-adverbials
(1) a. ?? È buio per due ore.
‘It is dark for two hours.’
PRESENTE
da-adverbials
(2) a. È buio da due ore.
‘It is dark since two hours.’
PRESENTE
b. ?? Era buio per due ore.
‘It was dark for two hours.’
IMPERFETTO
b. Era buio da due ore.
‘It was dark since two hours.’
IMPERFETTO
c. Fu buio per due ore.
‘It was dark for two hours.’
PASSATO REMOTO
c. ?? Fu buio da due ore.
‘It was dark since two hours.’
PASSATO REMOTO
d. È stato buio per due ore.
‘It is been dark for two hours.’
PASSATO PROSSIMO
d. ?? È stato buio da due ore.
‘It is been dark since two hours.’
PASSATO PROSSIMO
e. Era stato buio per due ore.
‘It was been dark for two hours.’
PIUCCHEPERFETTO
e. ?? Era stato buio da due ore.
‘It was been d. since two hours.’
PIUCCHEPERFETTO
As shown by (1a-c), per-adverbials combine felicitously with the
Passato Remoto but not with the Presente and the Imperfetto; on the contrary, as you see from (2a-c), da-adverbials combine with the Presente and
the Imperfetto but not with the Passato Remoto. Interestingly, the insertion
of the Perfect changes the pattern of acceptability. As shown by the perfect
sentences in (d) and (e), per modification is felicitous when the auxiliary
bears Presente and Imperfetto morphology but da modification is bad in
these cases. I will argue that some properties of Italian tenses are responsible for these facts.2
3. The analysis in short
In order to account for the data in (1) and (2) I will assume that
 Da-adverbials combine with temporal predicates to give temporally
homogeneous predicates
 Per-adverbials combine with temporal predicate to give temporally
non-homogeneous predicate
 Presente and Imperfetto require their complements to be temporally
homogeneous
 Passato Remoto requires its complements to be temporally nonhomogeneous
This explains the distribution in (a)-(c); moreover I will propose that
 the Perfect morphology in (d)-(e) is the spell-out of a semantic tense
Temporal homogeneity and the Italian Perfect
39
combining with a temporally non-homogeneous predicate.
This explains the distribution in (d)-(e).
4. Italian tenses
In order to implement this analysis, I will propose that the Italian temporal system has two lexical entries in its inventory of tenses: a tense selecting for temporally homogeneous predicates and a tense selecting for
temporally non-homogeneous predicates. The insight is that Imperfetto and
Presente are two forms of one and the same tense selecting for temporally
homogeneous predicates, while Passato Remoto and the Passato Prossimo
are two tense forms selecting for temporally non-homogeneous predicates.
This idea goes back to an observation of the ancient Latin grammarian Varro (116 B.C. - 27 B.C.), who argues (De Lingua Latina, X, 48) that we
should assume a basic division of the Latin tense forms into two stems, Infectum and Perfectum. According to Varro, while the Latin present and
past-imperfective verbal forms (like lego (I read-PRES) and legebam (I
read-PAST.IMP) ) are Infecta forms and are analogous to one another, the
Latin past-perfective forms (like legi (I read-PAST.PERF)) are Perfecta
forms and are to be opposed to the former (see Oldsjö 2001).
I will define the class of temporally homogeneous predicates as the class
of those predicates having the sub-interval property, the class of temporally
non-homogeneous predicates as the class of those predicates not having it
(Bennett and Partee 1978). Given a temporal property P of type <i, t>,
(3)
P is temporally homogeneous if i [P( i) i' [ i' i  P( i')]]
(4)
P is temporally non-homogeneous if i [P(i)i' [i’ i  P(i')]],
where i and i' are intervals
4.1.
Tense architecture
In order to develop a tense architecture appropriate to the proposal I am
making, I postulate that Italian tenses are to be decomposed into an homogeneity condition (which distinguishes Infectum the Perfectum) and a temporal variable (which distinguishes past from present)3. I will assume that
the homogeneity condition is the actual head of the tense projection whose
specifier position is filled by the time variable as represented by the trees
below:
40
Fabrizio Arosio
I. TEMPORA INFECTA
II. TEMPORA PERFECTA
Imperfetto
Passato Remoto
TP
TP
PAST i
PAST i
T'
T
HOM <it,it>
VP<it>
T'
T
N–HOM <it,it>
Presente
Reporters Present?
TP
TP
PRES i
PRES i
T'
T
HOM <it,it>
4.2.
VP<it>
VP <it>
T'
T
N–HOM <it,it>
VP<it>
Homogeneity conditions and temporal variables
Within this architecture, the head of the tense projection will be a predicate
restriction, namely, a function from predicates of time to predicates of
time, presupposing its complement to be temporally homogeneous or not.4
The lexical entries of the two Italian tenses will be
(5)
(6)
HOM =: Pt(P(t)): P is temporally homogeneous.
N-HOM =: Pt(P(t)): P is temporally non-homogeneous5
As you can see from the tense architecture above, the predicate restriction associated with the Tempora Infecta is HOM, the one associated
with the Tempora Perfecta is N-HOM. Therefore, in the case of the Tempora Infecta, the predicate restriction HOM takes a predicate of times as its
argument and it gives it back, if this latter is temporally homogeneous; in
the case of the Tempora Perfecta, N-HOM takes a predicate of times as its
argument and it gives it back, if it is temporally non-homogeneous. The
temporal predicate restrictions are what brings the contribution of tense into
the aspectual interpretation of a sentence, as we will see later. The obtained
property of times is, in turn, saturated by the denotation of the temporal
pronoun occupying the TP-Spec position.
As for the definition of the temporal pronoun, I will assume the straightforward analysis of Heim (1994), who developed the intuition that tenses
Temporal homogeneity and the Italian Perfect
41
are to be handled on a par with pronouns (Partee 1973; Enç 1986). Heim
observes that, just as a free instance of she can only refer to a female individual, a free instance of the temporal pronoun PASTi refers to a time before the time of the utterance, and therefore, since the contribution of gender has been treated as a presupposition (Cooper 1983; Heim 1982), we
should do the same with the temporal features associated with the temporal
variables. Temporal pronouns are thus defined as time variables carrying
presuppositions:
(7)
||PASTi||g,c = g(i) when g(i) < tc , undefined otherwise
(8)
||PRESi||g,c = g(i) when g(i) < tc , undefined otherwise
Definition (7) says that the denotation of PASTi is defined if the value that
the assignment function g assigns to the index i is a time before tc, the time
of the context of utterance; if it is defined, it is equal to the value that the
assignment function g assigns to the index i. Definition (8) says that the
denotation PRESi is defined if the value that the assignment function g assigns to the index i is a time not before tc; if it is defined, it is equal to the
value that the assignment function g assigns to the index i.
We can now see how the system works. Consider the following sentence:
(9)
Mario era
Mario was(Imperfetto)
‘Mario was sick.’
malato.
sick
The derivation of its LF will start with the state predicate Mario be sick. I
will distinguish event predicates from state predicates and I will assume
that the logical form of a state predicate has an explicit argument for times
as follows (Herweg 1991, Katz 1995)
(10) t(Mario be sick(t))
The state predicate is merged with the tense head and the obtained temporal
predicate is in turn filled by the temporal pronoun PASTi, as shown below
(11) PASTi (HOM( t(Mario be sick(t)) ))
42
Fabrizio Arosio
(12)

TP
PAST i
T'
T
HOM <it,it>
VP<it>
t( Mariobe sick(t))
Given the definedness conditions carried by the temporal variable and by
the predicate restriction HOM, we will have that (11) is defined if
(13) ||PASTi ||g,c is defined,
and if
(14) ||HOM(t(Mario be sick(t)))||g,c, is defined,
namely if g(i) < tc, and if t(Mario-be-sick(t)) is homogeneous, which is the
case, since state predicates are temporally homogeneous (Bennett and Partee 1978; Dowty 1979). If this is the case, then
(15) ||PASTi (HOM( t(Mario be sick(t)) ))||g,c = 1 iff
the time denoted by PASTi is a past time at which Mario is sick.
5. Durative per- and da-adverbials
I will assume durative per- and da-adverbials to be temporal modifiers
(Parsons, 1994), namely functions from time properties to time properties.
Intuitively, while per-adverbials take a temporal predicate and they give it
back by saying that it is true of an interval of a certain length (Dowty
1979), da-adverbials take a temporal predicate and, by introducing an extended now interval (Dowty 1979; McCoard 1978), they give back a predicate of times abutting an interval of a certain length of which the original
temporal predicate is true (von Stechow 2002; Musan 2000)6. The interesting property of these adverbs is that while per-adverbials quantize (Krifka,
1989) the temporal predicate they modify, da-adverbials make them homogeneous.
That per-adverbials turn all temporal predicates into quantized temporal
predicates (which are, a fortiori, non-homogeneous) is easy to see. I will
Temporal homogeneity and the Italian Perfect
43
assume that per-adverbials introduce universal quantification over times
(Dowty 1979).
(16) Per x time := Pt( t = x time long & t’(t’t P(t’))
The definition in (16) says that a per-adverbial takes a predicate of times P
as its argument and it gives back a temporal predicate denoting a set of
times i, which are x long, and for which every subinterval j of i belongs to
the set of times denoted by the original predicate P. More intuitively, the
adverb says that the predicate P is true of every subinterval of an interval
that is x long. These adverbs quantize the temporal predicate they modify,
namely they give a temporal predicate which, if true of an interval, is false
of every proper subinterval of that interval7.
(17) P is temporally quantized iff i[P(i)j [ji  P(j)]]
where i and j are intervals
Now, consider the time predicate John be sick for two days. As shown by
the picture below, if it is true of an interval i, it is false of every proper subinterval j of i trivially because j cannot be two days long if it is properly
included in i
John-be-sick-for-two-days
j2days
i = 2 days long
Da-adverbials turn all temporal predicates into homogeneous temporal
predicates. This is very easy to see as well. As proposed by von Stechow
(2002), these adverbials introduce an extended-now interval. I will assume
a slightly modified version of von Stechow's definition of durative seitadverbials, which works better for Italian
(18) da x tempo := PtI( t abuts I & P(It) & (I) = x time)
The definition in (18) says that a da-adverbial applies to a temporal predicate P and it gives back a temporal predicate denoting a set of times i such
that there is an x-time long interval I abutting i ("t abuts I" means that the
right bound of I is the left bound of t) and the set union of I and i belongs to
the set of times denoted by the original predicate P. Intuitively, it gives
back a property of times abutting an x-long interval of which the original
predicate is true. Consider now the temporal predicate in (20) obtained by
the application of the da-adverbial in (19) to the state predicate t(Mario be
44
Fabrizio Arosio
sick(t)):
(19) da due giorni (since two days):= PtI(t abuts I & P(It) &
(I)=2d.)
(20) tI( t abuts I & Mario-be-sick (It) & (I)=2 days)
The temporal predicate in (20) denotes a set of times i abutting an interval
of 2 days length I for which Mario is sick at i plus I, as shown by the picture below
time line
Mario's sickness
I = 2 days long
i
The "homogenizing" nature of da-adverbials is easy to understand. Consider the picture below
time line
Mario's sickness
I = 2 days long
j
i
If (20) is true of the interval i, it should be true of every subinterval j of i.
Since the duration measure two days is relative to the time I introduced by
existential quantification and not to i, and I abuts j, I  j will be always included in I  i. Therefore, if Mario is sick at I  i, he is sick at I  j, and
this shows that the obtained predicate is homogeneous.
While the quantizing character of for-adverbials explains why they do
not combine with a Tempus Infectum but do with a Tempus Perfectum, the
homogenizing nature of da-adverbials explains why they combine with a
Tempus Infectum, but not with a Tempus Perfectum. Consider the following sentences where a per-adverbial combines with the Passato Remoto but
not with the Imperfetto and the associated LFs:
(21) Mario fu
malato
Mario was(P. Remoto) sick
‘Mario was sick for two days.’
per
for
due
two
giorni.
days
Temporal homogeneity and the Italian Perfect

45
TP
PAST i
T'
T
N–HOM <it,it>
VP
perdue giorni
fortwo days
Pt (t = 2 days &
t’(t’t P(t’)))
(22) ?? Mario
‘Mario
era
was(Imperf.)

VP<it>
Mario essere malato
t( Mario be sick(t))
malato
sick
per
for
due
two
giorni.
days.’
TP
PAST i
T'
T
HOM <it,it>
VP
perdue giorni
fortwo days
Pt (t = 2 days &
t’(t’t P(t’)))
VP<it>
Mario essere malato
t( Mario be sick(t))
In both (21) and (22) we start the derivation of the LFs from the state
predicate t(Mario be sick(t)) and we obtain the non-homogeneous temporal predicate t( t = 2 days & t’(t’t  Mario be sick (t‘)) by formodification. In (21) this temporal predicate is merged with the tense head
N-HOM and in turn applied to the temporal variable PASTi. The definedness conditions for the obtained sentence require the denotation of PASTi to
be a past time and t( t = 2 days & t’(t’t  Mario be sick (t‘)) to be
non-homogeneous, which is the case. Therefore if PASTi refers to a time
before the time of the context of utterance, the sentence is felicitous when
uttered in that context, and it is true if and only if PASTi denotes a time
which is two days long and is characterized by Mario's sickness. On the
46
Fabrizio Arosio
contrary, in (22), the presupposition carried by HOM cannot be met, since
t( t = 2 days & t’(t’t  Mario be sick (t‘)) is temporally nonhomogeneous. This presuppositional-illness of (22) is then the root of its
ungrammaticality. As we observed before, a per- modified predicate is fine
under the Imperfetto and the Presente when receiving an habitual interpretation. This follows from the proposal I am making. Since habits are temporally homogeneous, habitual predicates can meet the presupposition associated with a HOM head. I will propose habitual predicates to be obtained by
application of an HABITUAL operator which stativizes the predicate it
modifies. For further discussion see Scheiner's (2002) MA thesis.
On the other hand, the homogeneous character of a temporal predicate
obtained by da-modification allows it to combine with the Imperfetto for
the opposite reason but prevents it from combine with the Passato Remoto,
as shown by the sentence (23) and its LF:
(23) ?? Mario
Mario
fu
malato
was-P.Remoto sick

da
due
since two
giorni.
days
TP
PAST i
T'
T
N–HOM <it,it>
VP
da du egio rn i
sincetwo da ys
PtI (t ab uts I & P(It)
& (I) = 2 da ys)
VP<it>
Ma rio essere malato
t(Ma rio be sick(t))
As we see from its LF, sentence (23) suffers from presuppositional-illness,
since the non-homogeneity presupposition carried by the N-HOM head
cannot be met by the temporal property tI(t abuts I & Mario-be-sick
(It) & (I) = 2 days), which is homogeneous. Clearly, such a temporal
property combines with a Tempus Infectum, since it meets the homogeneity
presupposition carried by the HOM head.
State predicates combine however with the Passato Remoto without any
overt for-adverbial modification, as shown by the sentence below:
(24) Mario fu malato (*ed è ancora malato).
Temporal homogeneity and the Italian Perfect
47
‘Mario was-PassatoRemoto sick (*and he is still sick).’
Sentence (24) describes some sickness of Mario’s which is terminated in
the past, as shown by the non-availability of the continuation in brackets. I
assume that, in such cases, in order to fulfill the non-homogeneity presupposition associated with the Passato Remoto, the temporally homogeneous
state predicate is maximalized (and thus temporally quantized) by the covert application of the following maximality operator
(25) MAX:= Pt(P(t) & t'(tt' & P(t'))
The application of MAX gives us the following correct LF for (24)
(26) PASTi(N-HOM(t(Mario-be-sick(t)&t'(tt' & Mario-be-sick (t')))).
According to (26), the application of the MAX operator correctly gives
us that Mario's sickness is terminated in the past. On the one hand the application of the MAX operator gives the same quantizing effect as the application of for adverbials, but on the other hand, this latter is different from
the former since it does not entail terminativity. In fact, nothing prevents
Mario's sickness to hold up to the speech time given the LF associated with
(21). According to my intuitions, this seems to be a an unwelcome prediction since I understand Mario's sickness to be terminated in the past in (21).
I found however a significant geographical variation on the judgements
and, moreover, if we have a brief look at available corpus data we do find
sentences like:
(27)
La cupola del Brunelleschi fu per lungo tempo, ed è ancora, il simbolo visivo dell'intera città.
‘The cupola from Brunelleschi was-Pass.Remoto for long time, and
it is still, the symbol visual of-the whole city.’
which require the for-modified states not to be necessarily terminated before the speech time. I argue that we find a terminativity effect due to
pragmatic factors associated with for-adverbials: since for-adverbials are
downward entailing (John ran in the park for five minutes entails John ran
in the park for four minutes), in order to be maximally informative, the
length indicated by the for-adverbial should be the length of the maximal
interval for which the predicate is true. This explains why we intend that
Mario's sickness is terminated in the past in (21).
48
Fabrizio Arosio
6. Italian Perfect I: a special tense
As we have seen in (1) and (2), the presence of the Perfect changes the distribution of the adverbials under the same tense morphology (borne by the
auxiliary verb). In recent work, von Stechow (2001) (see also Paslawska
and Stechow (2003) in this volume) argued that the Passato Prossimo morphology in (d) is a variant of the Passato Remoto inflection and he suggested that the same temporal meaning can be spelled out by the Passato Remoto or by the Passato Prossimo in Italian. While, on the one hand, this proposal correctly predicts the adverbial distribution in (d), on the other, it reveals its weakness if we consider the contrast below:
(28) Maria
Maria
vissuto
lived
sposerà
un
marry-3singFUT a
uomo che ha
man who have-3singPRES
a NY.
in NY.
‘Maria will marry a man who has lived in NY.’
(29) Maria
Maria
sposerà
un
marry-3singFUT a
visse
live-3sing.PassatoRemoto
uomo che
man who
a NY.
in NY
‘Maria will marry a man who lived in NY.’8
Sentence (28) is temporally ambiguous. It can mean: (i) that Mary will marry a man in the future with respect to the speech time and that this man
lives in NY in the past with respect to the speech time; or (ii) that Mary will
marry a man in the future with respect to the speech time and that this man
lives in NY in the past with respect to the future marrying event (thus his
living in NY may be in the future with respect to the speech time), as represented by (30a-b) below
(30) a.
living-in-NY
speech time
marrying
Temporal homogeneity and the Italian Perfect
49
b.
speech time
living-in-NY
marrying
Sentence (29) is not temporally ambiguous. According to (29), the livingin-NY can only be in the past with respect to the speech time, namely: (29)
can only have the temporal interpretation represented in (30a). If we consider the Passato Remoto and the Passato Prossimo to be free spell-out variants of one and the same tense, we have to stipulate that the Passato Prossimo is semantically ambiguous in order to account for the meaning variations of (28), so, we have to assume that under the reading (30b) the
Passato Prossimo is the spell-out of something else. We can easily get rid
of this ambiguity if we assume that the perfect morphology in (28) is the
spell-out of a Priorean (1957) temporal operator in the scope of the present
tense as represented below
(31)
TP
ti
T'
T
<it,it>
PERFECT
VP<it>
(32) PERFECT:= Ptt'[t'<t & P (t')] : P is temporally non-homogeneous.
The definition in (32) says that the perfect takes a time predicate as its argument and gives a predicate of the times which are after a certain time at
which the original temporal predicate holds, and it presupposes this latter to
be temporally non-homogeneous. As observed by von Stechow (2002), the
Perfect as defined in (32) can be seen as the object language representation
of the truth conditions of the Priorean past operator, since it introduces an
existential quantification over times preceding the time denoted by tense.
By introducing existential quantification, the Perfect turns all the predicates
it modifies into temporally homogeneous predicates. Let us illustrate this.
Consider the following temporal property obtained by Perfect modification
from the temporally quantized predicate Mario be sick for two days
(33) tt' [t' < t & Mario be sick for two days (t')]
As shown by the picture below, if (33) is true of a interval i, it is true of
50
Fabrizio Arosio
every subinterval j of i, trivially because an interval which is before another
interval is before every subinterval of that interval
time line
j
Mario be sick for two days
i
According to this proposal, the LF associated with (1d) will therefore be:
(34)

TP
PRES i
T'
T
HOM <it,it>
PERFECT
Ptt' [t' < t & P (t')]:
P is temporally
no n–ho mogeneou s
VP<it>
perdu eore
fortwoho urs
Pt (t = 2 ho urs &
t’(t’t P(t'))
VP<it>
it to be da rk
t(it to be da rk (t))
In this analysis, the difference between the Passato Remoto and the Piuccheperfetto (the Italian Pluperfect form you find in (1e) and in (2e)) will
depend on whether we find a past or a present temporal pronoun in SpecTP position. The ungrammaticality of (2d-e) follows from the nonhomogeneity presupposition we associated with the Perfect in (32). Since
the Perfect presupposes its complement to be non-homogeneous, it cannot
combine with a temporal predicate obtained via da-adverbial; this nonhomogeneity presupposition is moreover responsible for the fact that eventualities described under the Perfect get a terminative interpretation in Italian.
One weak point of this proposal is that our durative temporal adverbials
should always occur in the scope of the perfect. This seems to me a quite
strong stipulation and I think it is a sign that our description of the Italian
perfect is not yet entirely correct. In our analysis, we have so far described
Temporal homogeneity and the Italian Perfect
51
the perfect as a temporal operator in the scope of tense. I think that this is
wrong and that the scope stipulations are a indication of this error: the perfect is a tense. In order to develop this intuition, I propose the Passato Prossimo and the Piuccheperfetto to be two forms of a special tense, the perfect.
In harmony with the architecture of tense proposed in this paper, I assume
the perfect to be a special predicate restriction as defined in (32), namely a
special tense head selecting for temporally non-homogeneous predicates,
whose specifier position is filled by a past or a present time variable as represented below9
ITALIAN PERFECT I
Passato Prossimo
Piuccheperfetto
TP
PRES i
TP
T'
T
PERFECT
PAST i
VP<it>
T'
T
PERFECT
VP<it>
This analysis gives us the correct scope order: temporal adverbials are
always in the scope of the perfect with no further stipulations.
7. Aspect and event predicates
Until now, we have discussed state predicates which we assumed to denote
properties of times. But, as well known (Davidson 1976), we need events in
our ontology. Following Herweg (1991) and Katz (1995), I distinguish between state and event predicates by assuming that, while a state predicate
has an explicit argument for times, an event predicate has an explicit argument for events as below
(35) || to love || := yxt(love(t),(x),(y))
(36) || to sleep || := xe(sleep(e),(x))
As will be clear from (35) and (36), a temporal predicate representing a
state belongs to the logical type <i, t> and can be modified by a durative
temporal adverbial or be saturated by the tense variable, while an event
predicate belongs to the type <e, t > and needs to be type shifted to undergo
temporal modification. I assume that, in the LF of an event sentence, implicit aspectual operators are responsible for this type shifting by localizing
52
Fabrizio Arosio
the described event with respect to time (Klein 1994; Kratzer 1998;
Stechow 2001; Musan 2000). The common aspectual operators are the inclusion operator, and its complementary. The inclusion operator, also called
Perfective operator, since it is responsible for the perfective reading of an
event sentence, localizes the described event within a time: it takes a property of events and it gives a property of times including the temporal trace
of the event. Its complementary, also called Imperfective operator, since it
is responsible for the imperfective reading of an event sentence, localizes
the described event as surrounding a time: this operator takes a property of
events and it gives a property of times properly included in the temporal
trace of the event. The Perfective and Imperfective aspectual operators are
defined as follows (where  is a function taking an event and giving its
temporal trace.10):
(37) Perfective =: ||PFV|| = Pte(t  (e) & P(e))
(38) Imperfective =: ||IPV|| = Pte((e) t & P(e))
Interestingly, a temporal property formed via the IPV operator will always
be homogeneous, as shown by the picture below:
||(e)|| = i
j
time line
k
If j is included in the temporal trace of e, every subinterval k of j will be
included as well: therefore we expect a temporal property formed via the
IPV operator to combine with a Tempus Infectum. On the other hand, a
temporal property formed via the PFV operator will always be nonhomogeneous, therefore we expect it to combine with a Tempus Perfectum
||(e)|| = i
j
time line
k
As shown above, if j includes the temporal trace of e, it is not the case that
every subinterval k of j will be included in j. Consider now the following
sentence describing an event under the Passato Remoto and its associated
LF:
(39) Mario mangiò
Mario eat-P.REMOTO
‘Mary ate an apple.’
una
an
mela.
apple
Temporal homogeneity and the Italian Perfect
53
(40)

TP
PAST i
T'
T
N–HOM <it,it>
AspP<et,it>
PFV:
Pte(t  (e)
& P(e))
VP<et>
Ma rio mang iareun amela
e( Ma rioeat an ap ple(e))
The temporal predicate obtained via the PFV operator meets the non homogeneity condition carried by the N-HOM head. This explains why event
sentences always have terminative interpretations under the Passato Remoto, as shown by the bad continuation of the sentence in (41):
(41) Mario mangiò una mela, *e la sta ancora mangiando.
‘Mario eat-P.REMOTO an apple, *and it he-is still eating.’
Analogously, a temporal predicate obtained via the PFV operator, but not
one obtained via the IPV operator, meets the non-homogeneity presupposition associated with the PERFECT tense head. This correctly predicts that
events predicates described under the Perfect get a terminative interpretation as shown by the bad continuation in (42):
(42) Mario ha mangiato una mela, *e la sta ancora mangiando.
‘Mario has eaten an apple, *and it he-is still eating.’
Going back to state predicates, we find that they are ambiguous between
a terminative and an inchoative interpretation under the Passato Remoto.
Consider the following sentence:
(43) Fu
buio.
(it) was-P.Remoto dark
As we have already seen (see (24)), sentence (43) can mean that there is
some past terminated darkness state, but it can also mean that it became
dark in the past. Given the non-homogeneity condition associated with the
54
Fabrizio Arosio
Passato Remoto, these two readings are the result of shifting a homogeneous predicate into a non-homogeneous one in order to satisfy it. As we have
seen, the terminative reading is obtained via maximalization of the state
predicate. The inchoative reading of (43) I propose to be obtained by the
application of the covert "achievementizing" operator below
(44) BECOME =: Pe (becomee(P)). The event e is a becoming with
result P, where P is a state.
As shown by (45), the predicate obtained via the become operator is in
turns perfectivized and selected by the Passato Remoto
(45) a.

TP
PAST i
T'
T
N–HOM <it,it>
AspP<et,it>
PFV:
Pte(t  (e)
& P(e))
<it,et>
BECOME = :
Pe (become e(P))
VP<it>
t(it– be– da rk (t))
b. PASTi (N-HOM (te(t  (e) & becomee(It-be-dark))))
Under this reading, the sentence is true if and only if g(i) is a past time including the time of a becoming dark event. Interestingly, the analysis predicts that the result state can hold up to speech time since the embedded
state of result is not temporally constrained, as shown by the sentence below (Bertinetto 2001):
(46) La sua squadra preferita aveva perso. Gianni ne ebbe un forte mal di
pancia che ancora non gli è passato.
‘His preferred team had lost. Because of this, Gianni got a belly
ache, which is still paining him.’
Temporal homogeneity and the Italian Perfect
55
8. Italian Perfect II: telic predicates and Perfect
It is however not true that da-adverbials do not combine with the Perfect at
all. In fact they do when the predicate is a change of state predicate:
(47) Il Parco Disney Studios ha aperto i cancelli da sei ore.
‘The Park Disney Studios has opened the gates since six hours.’
Sentence (47) says that the Disney Park opened its gates, and that the
gates have been open for six hours (up to now); in (47), the adverbial modifies the state brought about by the completion of the opening event and it
says that this state holds for a six-hours-long interval abutting the speech
time.
I argue that, in order to give a correct analysis of sentences like (47), we
have to assume that, in Italian, the perfect morphology is ambiguous between the spell-out of a special Tempus Perfectum, as defined in section 6,
and the spell-out of a result state construction. The perfect morphology in
(47) is the spell-out of this latter: it introduces a state of result brought
about by the completion of a telic event which can be modified by temporal
adverbials or saturated by tense.
That the perfect is ambiguous between these two meanings is clear if we
consider the contrast below:
(48) a. Alle tre, il Parco Disney Studios ha aperto i cancelli da 6 ore.
‘At three o'clock, the Park Disney Studios has opened the gates
since six hours.’
b. Alle tre, il Parco Disney Studios ha aperto i cancelli per sei ore.
‘At three o'clock, the Park Disney Studios has opened the gates
for six hours.’
In (48a-b) da- and per-adverbials11 modify the result state brought about by
the completion of the opening event but, interestingly, in (48b) the adverb
alle tre (at three o'clock) localizes the opening event, while in (48a) it localizes the right bound of the result state (the being open of the gates). This is
a clear indication of the fact that the perfect conveys two different meanings in (48a-b): in (48a) it localizes the at-three-o'clock opening event in
the past with respect to speech time, while in (48b) it introduces the state of
the gates being open.
One might object that this does not prove anything, since these facts
could be explained by assuming that the perfect morphology in (47) is the
spell-out of a temporal operator the post state of the opening event, namely
56
Fabrizio Arosio
the forever holding state of the opening event having culminated, which
Parsons (1990) calls resultant state, and that this post state is what is modified by the da-adverbial. This explanation cannot be correct if we consider
the unavailability of the continuation of (47) represented below:
(49) ?? Il Parco Disney Studios ha aperto i cancelli da sei ore ma la
polizia li ha chiusi un'ora fa.
‘The Park Disney Studios has opened the gates since six hours,
but the police has closed them one hour ago.’
Sentence (49) shows that the da-adverbial does not modify the post state of
the opening event but rather its result state, since the gates should be open
at speech time.
Recently, it has been argued that there are two classes of telic predicates, those that have an accessible result state, e.g. to open the gates, and
those that do not, e.g. to send a letter (von Stechow 2002, Kratzer 2000).
According to this analysis, only the former should occur in perfect of result
constructions. I argue that this cannot be correct if we consider an Italian
sentence like:
(50) Mario ha spedito il suo articolo da una settimana.
‘Mario has sent his paper since one week.’
which, according to von Stechow and Kratzer should not convey a result
state meaning. But as we have seen that da-adverbials do modify result
states, the predicate spedire un articolo (to send a paper) do have an accessible result state. On the other hand, if we say that both spedire un articolo
(to send a paper) and aprire i cancelli (to open the gates) have an accessible
result state, we have to say why for-adverbials can modify the former but
not this latter, as shown by the contrast below:
(51) ?? Mario ha spedito il suo articolo per una settimana.
‘Mario has sent his paper for one week.’
(52) Il Parco Disney Studios ha aperto i cancelli per sei ore.
‘The Park Disney Studios has opened the gates for six hours.’
In order to explain this fact, I argue that per-adverbials (for-) presuppose
the result state they modify to be reversible
(53) State reversibility:
S is a reversible state iff t(S(t)  t'(t<t' & S(t')))
This reversibility presupposition associated with per-adverbials prevents
Temporal homogeneity and the Italian Perfect
57
them from combining with non reversible result states like to be dead, or
to be sent.
In order to give an analysis for (47), we have to formulate a definition of
change of state predicates which gives us the correct result when occurring
in result state constructions. In recent work, Kratzer (2000) and von
Stechow (2002) have proposed two alternative but similar definitions of
accomplishment predicates sensitive to result state modification. While in
Kratzer's analysis transformative predicates denote relations between events
and individual states having a certain property, in von Stechow's they denote relations between events and properties of states, in order to account
for the different meanings of adverbs modifying causative verbs (see also
von Stechow 2000). Both Kratzer and von Stechow assume then that special operators called "aktionsart choosers" convert this relation into an
event predicate or into a state of result predicate. Given that in my system I
assumed states not to be primitive entities but state predicates to denote
property of times, if we follow Kratzer's proposal we come up with the
counterintuitive result that transformative predicates denote a causative relation between events and times having a certain property as defined below
(54) to open := xet(CAUSEw(e, t) & open(x)(t))
The only way of making sense of this is to assume Lewis's idea that times
are event slices of a world; we shall avoid going into such a discussion
which would be otiose to the aim of this paper. On the other hand, if we
follow von Stechow's proposal and we reformulate his definition of transformative verbs in temporal terms, we obtain the following definition
which integrates more naturally into the proposal I am making, without
touching any ontological question about the nature of time
(55) to open := xyeP(CAUSEw(e,P) & Become(e,P) & Agent(e,y) &
P = open x), where P is a state
However, (55) does not give us the right results when we apply to it the
“aktionsart choosers” described in (57) and (58) below and we try to calculate the truth condition of the “CAUSE” proposition. This because, according to (55), transformative verbs denote causal relations between events and
sets of times. Chierchia (p.c)12 suggested to me that a possible way of representing transformative predicates in the system I am proposing in this
paper is to assume them to denote complex relations between an event and
a time which abuts its temporal trace and of which the introduced result
state holds, as described below13
(56) to open:= xyet(t><(e) & CAUSEw(e,
open(x)(t)) &
58
Fabrizio Arosio
Agent(e,y))
According to Kratzer's and von Stechow’s proposals, the relation in (56)
is converted into an event predicate or into a result state predicate by the
application of the following "aktionsart choosers", which I redefined in
temporal terms
AKTIONSART CHOOSERS -FIRST TRY
(57) RESULT:= Rte(R(e,t))
(58) EVENT := Ret(R(e,t))
According to (58), the eventive aktionsart chooser closes the temporal variable t and gives us a property of events, which can in turn enter the derivation of the LF of an eventive sentence. On the other hand, according to
(57), the RESULT aktionsart chooser closes the event variable e and gives
us a property of times which are in a complex relation with e. Unfortunately, the result state predicate obtained by applying (57) to (55), as represented below
(59) te(t><(e) &CAUSEw(e, the-gates-be-open(t)) & Agent(e,The-PD-Studios) )
cannot meet the predicate restrictions associated to the tempora Infecta
since is not temporally homogeneous. Its non-homogenous nature depends
on the “t><(e)” condition; as we can see from the picture below, it is not
true that every subinterval k of an interval j abutting the temporal trace of e
abuts the temporal trace of e
(60)
(e)
j
time line
k
This would prevent a result state predicate to combine with a Tempus Infectum. This abutting condition is a problem for von Stechow’s (2002)
formulation as well since there we find a BECOME operator. Von Stechow
assumes, in fact, that result state predicates can be modified by seitadverbials (the analogous of Italian da-) in German and he correctly points
out that positional seit-adverbials in German presuppose their argument to
be temporally homogeneous as shown by the contrast below:
(61) Dieter ist seit 1975 in Düsseldorf. (von Stechow 2002)
‘Dieter is since 1975 in Düsseldorf.’
Temporal homogeneity and the Italian Perfect
59
(62) * Dieter ist seit 1975 achtmal in Tübingen.
‘Dieter is since 1975 eight-times in Tübingen.’
Since (59), and the analogous result state predicate in von Stechow's analysis, is not temporally homogeneous it cannot be modified by seitadverbials. Moreover we find one more problem in von Stechow's (2002)
proposal given his assumption that the Present tense denotes the speech
time conceived as a point in German. Since a result state property holds of
an interval abutting the temporal trace of the becoming event, when it combines with the present tense, the obtained temporal proposition will be true
if the speech time instant abuts the temporal trace e. A quite strong result,
which requires a sentence like:
(63) Ich habe meine Brille verloren.
‘I have lost my glasses.’
to be true if the loosing event abuts the time at which the sentence is uttered.
In order to solve these problems I assume the right definition of the result operator to be
(64) RESULT := RteI(R(e, I) & t  I)
OFFICIAL
According to (64), the result operator is a stativizer and its application to
(55) gives us the correct definition of result state predicate below
(65) teI (I ><(e) & CAUSEw(e,
Agent(e,The-P-D-Studios) & t  I )
the-gates-be-open(I))
&
This temporal property is homogeneous and meets the homogeneity
condition associated by a da-adverbials, as observed by von Stechow, and
the presuppositions introduced by the predicate restrictions associated with
the temporal Infecta. The correct LF for (47) will be therefore
60
Fabrizio Arosio
(66) a.
TP
PRES i
T'
T
HOM <it,it>
VP <it>
1
VP<et,it<i,t>>
2
VP<<<et >,<it>>,t>
3
1 = da due giorni, since two days. PtI( t abuts I & P(It) &
(I)=2 days)
2 = RteI(R(e, J) & t  J)
3 = et(t><(e) & CAUSEw(e, open-the gates(t)) &
Agent(e,D.S.))
b. tI( t abuts I & eJ (J><(e) & CAUSEw(e, the-gates-beopen(J)) & Agent(e,The-P-D-Studios) & (I PRESi)J & (I)=2
days)
Given the LF in (66), there is one more thing that we should explain.
According to how (47) is intended, the 2-day long interval introduced by
the da-adverbial is understood as left abutting the temporal trace of the
opening event and right abutting the speech time. This does not follow
from the application of the da-adverbial to (65). I argue that this left abutting effect is due to pragmatic factors. Since da-adverbials sentences are
downward entailing (Mario e` malato da due ore (Mario is sick since two
hours) entails Mario e` malato da un'ora (Mario is sick since one hour)), in
order to be maximally informative, the length indicated by the da-adverbial
should be the length of the maximal interval for which we have evidence
that the predicate is true. This explains the left abutting effect of daadverbial modification of result state predicates.
9. Generalizing across Romance languages
Da- and per-adverbials follow the same pattern of distribution in Romance
languages.
FRENCH
Consider the following facts in French, where pendant- and depuis-
Temporal homogeneity and the Italian Perfect
61
adverbials are the analogues of the Italian da- and per-adverbials:
Pendant-adverbials
(67a) ??La fenêtre est sale pendant deux jours
The window is-PRÉS dirty for two days
[Présent]
(b) ??La fenêtre était sale pendant deux jours
The window was-IMP dirty for two days
[Imparfait]
(c) La fenêtre fut sale pendant deux jours
The window was-PS dirty for two days
[Passé Simple]
Depuis-adverbials
(68a) La fenêtre est sale depuis deux jours
The window is-PRÉS dirty since two days
[Présent]
(b) La fenêtre était sale depuis deux jours
The window was-IMP dirty since two days
[Imparfait]
(c) ??La fenêtre fut sale depuis deux jours
The window was-PS dirty since two days
[Passé Simple]
As we can see from (67) and (68), pendant-adverbials combine with the
Passé Simple but not with Présent and Imparfait while depuis-adverbs
combine with the Présent and the Imparfait but do not with the Passé Simple. Exactly as in Italian.
ROMANIAN
In Romanian, where the analogues of per-adverbials is the bare durative
adverbials14 and the analogous of da-adverbials the de-adverbials, we find
the same distribution as well:
Bare durative-adverbials
(69a) ?? Ion e bolnav mult timp
Ion is-PRES sick long time
[Prezentul]
(b) ?? Ion era bolnav mult timp
Ion was-IMP sick long time
[Imperfectul]
(c) Ion fusse bolnav mult timp
Ion was-P.SIMP sick long time
[Perfectul Simplu]
62
Fabrizio Arosio
De-adverbials
(70a) Ion e bolnav de 2 zile
Ion is-PRES sick since 2 days
[Prezentul]
(b) Ion era bolnav de 2 zile
Ion was-IMP sick since 2 days
[Imperfectul]
(c) ?? Ion fusse bolnav de 2 zile
Ion was-P.SIMP sick since 2 days
[Perfectul Simplu]
Moreover, Romanian, French and Italian pattern in the “Perfect” distribution as shown below:
Bare /Pendant /Per
adverbials
ROMANIAN
(71)a. Ion a fost bolnav mult timp
Ion has been-PC sick long time
De/Depuis /Da
adverbials
b. ??Ion a fost balnav de 2 zile
Ion has been-PC sick since 2 days
FRENCH
(72)a. Ion a été malade pendant longtemps b.??Ion a été malade depuis longtemps
Ion has been-PP sick for long-time
Ion has been-PP sick since longtime
ITALIAN
(73)a. Ion è stato malato per molto tempo
Ion is been-PP sick for long time
b. ??Ion è stato malato da due giorni
Ion is been sick since 2 days
Given the facts above, it is plausible to claim that we should extend our
analysis of Italian tenses to most Romance languages.
10. Conclusions
In this paper I have shown how temporal homogeneity accounts for the distribution of durative adverbials across different tense forms and for how the
progressive and habitual readings appear under the same tense forms; I
propose a tense decomposition in which tense itself has some influence in
the aspectual interpretation of a sentence. Moreover I have shown that the
Italian Perfect morphology is ambiguous in at least two ways by being the
spell-out first of a state of result construction and next of a special tense. I
additionally propose the generalizability of this analysis across Romance
languages; further cross-linguistic comparison with different and unrelated
Temporal homogeneity and the Italian Perfect
63
languages will no doubt provide further evidence for the homogeneity account I proposed.
Notes
* I wish to thank Arnim von Stechow and Graham Katz who helped me in finding a coherent story behind my foggy intuitions and Sam Featherston for English corrections. This work also benefited from conversations with Andrea
Bonomi, Gennaro Chierchia, Uli Sauerland and Winnie Lechner. Of course, only I am responsible for errors and failings. This research is part of the SFB441
activity.
1. For a discussion of the behavior of such adverbs in other languages see the articles of Fassi, Rathert, Giannakidou, Musan and Iatridou et al. in this volume.
2. (1a-b) are perfectly fine under a habitual interpretation in an appropriate context. As we will see later, this is predicted by the proposal I am making in this
paper.
3. We shall not discuss the case of future tenses in this paper since they do not add
any additional relevant data to our discussion. They could be characterised as
tenses having temporal variables only defined for future times.
4. A presupposition associated with a sentence narrows down the set of possible
contexts in which the sentence can be felicitously uttered (see Stalnaker 1978;
Lewis 1979; Heim 1982).
5. The denotation of the application of the predicate restriction to its complement
is therefore
- ||HOM (P) ||g,c = P when P is temporally homogeneous, undefined otherwise.
- ||N-HOM (P) ||g,c = P when P is temporally not homogeneous, undefined otherwise.
6. As for the German SEIT-POSITION-ADVERBIAL described by Musan
(2002), the Italian da- doesn't seem to have an English counterpart. Peradverbial is the Italian counterpart of the English for-.
7. Notice that being quantized is different from being non-homogeneous. See negation scope in (4) and (17).
8. Actually, this is not a good translation for the Italian (29) since it is an ambiguous English sentence which can have both the (30a) and (30b) readings.
9. As in § 4.1, I will ignore the case of the future. A future time variable can clearly fill the SpecTP position of the perfect tense head and this combination is
what is spelled out by the Futuro Anteriore morphology in Italian.
10. Clearly in the definition (38) we do not consider the modal aspect of the Imperfective aspect discussed in Bonomi (1999) since it goes beyond the purpose of the present work.
11. Vide Piñón (1999) and von Stechow (2000) for a discussion about result state
modification of for-adverbials.
64
Fabrizio Arosio
12. Of course, I am the only responsible for mistakes and errors.
13. The definition in (56) does not follow von Stechow’s idea that transformative
verbs denote a relation between an event and a property of a state, i.e. something propositional, but it is more Kratzer’s, since it assumes these verbs to
denote relations between events and individual times. Von Stechow (2000)
argues that we need the property and not the individual state in order to account for the restitutive readings of result state sentences modified by expressions like again/wieder/di nuovo. Chierchia (pc), again, suggested to me that
given (56) and the aktionsart choosers below we could easily account for the
repetitive reading of di nuovo/again modification by assuming these modifiers to denote a partial function from sets of object into sets of object as below
(i) di nuovo/again (P) = P when | P | ≥ 2, not defined otherwise
In this case, the restitutive reading is no more obtained by modification of the
state property, but always as a modification of a individual property (time
property) which enters in a CAUSE relation with a certain implicitly given
event. Again, I am the only responsible for mistakes and errors I made out of
Chierchia’s intuitions.
14. The prepositional pentru-adverbials (for-adverbials) are result state modifiers
in Romanian.
References
Bennett, Michael, and Barbara Partee
1978
Toward the Logic of Tense and Aspect in English. Bloomington, Indiana: Indiana University Linguistics Club.
Berrettoni, P.
1972
La Metafora aspettuale. Studi e Saggi Linguistici 12: 250-59.
Bertinetto, Pier Marco
1986
Tempo, Aspetto e Azione nel verbo italiano: il sistema dell' indicativo. Firenze: Accademia Della Crusca.
2001
On a Frequent Misunderstanding in the Temporal-Aspectual Domain: The 'Perfective-Telic' Confusion. In Semantic Interfaces [Reference, Anaphora, Aspect], Bruno Cecchetto, Gennaro Chierchia
and Maria Teresa Guasti (eds.), 177-210. Stanford: CSLI.
Bonomi, Andrea
1997
Aspect, quantification and when-clauses in Italian. Linguistics and
Philosophy 20: 469-514.
1999
Semantical remarks on the progressive readings of the imperfective.
Milano, Manuscript.
Cooper, Robin
1983
Quantification and Syntactic Theory. Dordrecht: Reidel.
Temporal homogeneity and the Italian Perfect
65
Davidson, Donald
1976
The logical form of action sentences. In The Logic of Decision and
Action, Rescher, N. (ed.). Pittsburgh: University of Pittsburgh Press.
Devoto, G.
1940
L'«aspetto» del verbo. Lingua Nostra, 2: 35-38.
Dowty, David
1979
Word Meaning and Montague Grammar. Dordrecht: Reidel.
Enç, Mürvet
1986
Toward a Referential Analysis of Temporal Expressions. Linguistics
and Philosophy 9.1: 405-426.
Giorgi, Alessandra, and Pianesi, Fabio
1997
Tense and Aspect: From semantics to morphosyntax. New York: Oxford University Press.
2001
Ways of Terminating. In Semantic Interfaces [Reference, Anaphora
and Aspect], Carlo Cecchetto, Gennaro Chierchia and Maria Teresa
Guasti (eds.), 211-277. Stanford, California: CSLI.
Herweg, Michael
1991
Perfective and imperfective aspect and the theory of events and
states. Linguistics 29: 969-1010.
Heim, Irene
1982
The semantics of definite and indefinite noun phrases. Ph.D. dissertation, University of Massachusetts, Amherst.
1994
Comments on Abusch's theory of tense, MIT manuscript.
Katz, Graham
1995
Stativity, genericity, and temporal reference. Ph.D. dissertation, University of Rochester.
Klein, Wolfgang
1994
Time in Language. New York: Routledge.
Kratzer, Angelika
1998
More Structural Analogies Between Pronouns and Tenses. Paper
presented at SALT 8, Cambridge, Mass.
2000
Building Statives: Berkeley Linguistic Society.
Krifka, Manfred
1989
Nominalreferenz und Zeitkonstitution. München: Wilhelm Fink.
Lewis, David
1973
Counterfactuals. Harvard University Press.
1979
Scorekeeping in a language game. Journal of Philosophical Language 8: 339-359.
Lucchesi, V.
1971
Fra grammatica e vocabolario: Studio sull'«aspetto» del verbo italiano, Studi di Grammatica Italiana 1, Firenze: Sansoni.
66
Fabrizio Arosio
McCoard, Robert W.
1978
The English Perfect: Tense-Choice and Pragmatic Inferences. Amsterdam: North-Holland.
Musan, Renate
2000
The Semantics of Perfect Constructions and Temporal Adverbials in
German. Humboldt Universität: Habilitationsschrift.
2002
The German Perfect: Its Semantic Composition and Its Interactions
With Temporal Adverbials. Kluwer Academic Publishers.
Oldsjö, F.
2001
Tense and Aspect in Caesar's Narrative. Ph.D. dissertation, Uppsala
Universitet.
Parsons, Terence
1994
Events in the Semantics of English: A Study in Subatomic Semantics.
Cambridge: MIT Press.
Partee, Barbara
1973
Some Structural Analogies between Tenses and Pronouns in English.
Journal of Philosophy 70: 601-609.
Paslawska, Alla, and Arnim von Stechow
(this volume) Perfect Readings in Russian
Piñón, Christopher
1999
Durative adverbials for result states. Universität Düsseldorf: Manucript. Available under: www.phil-fak.uni-duesseldorf.de/~pinon/dars.html.
Prior A.
1957
Time and Modality. Oxford: Clarendon Press.
Scheiner, Magdalena
2002
Temporale Verankerung habitueller Sätze. Magisterarbeit, Universität Wien.
Stalnaker, Robert C.
1978
Assertion. In Syntax and Semantics Vol. 9: Pragmatics, P. Cole
(ed.). Academic Press, New York.
Stechow, Arnim von
2000
How are Results Represented and Modified? Remarks on Jäger &
Blutner's Anti-Decomposition. In Adverbs: Proceedings of the Oslo
Conference on Adverbs, Fabricius-Hansen, Cathrine, Ewald Lang,
and Claudia Maienborn (eds.).
2001
The Janus Face of Aspect. Talk given at the Universita` degli Studi
di Milano, October 5th 2001.
2002
German seit ‘since’ and the ambiguity of the German Perfect. In
More than Words: A Festschrift for Dieter Wunderlich, Barbara
Stiebels and Ingrid Kaufmann (eds.), 393-432. Berlin: Akademie
Verlag.
Temporal homogeneity and the Italian Perfect
67
Varro, Marcus Terentius
1996
De Lingua Latina. In De Lingua Latina X. A New Critical Text and
English Translation With Prolegomena And Commentary, Daniel
Taylor (ed.). Amsterdam: Benjamins.
68
Fabrizio Arosio
D
durative adverbs
da ('since') 42
da 'since' 37
in Romance languages 60
per ('for') 42
Perfect
in Italian 48
Piuccheperfetto 50
Presente 39
Q
quantized predicates 43
H
Herweg, Michael 41
S
Stechow, Arnim von 43
I
Imperfetto 39
M
T
tempora infectum 39
tempora perfectum 39
terminativity 47
maximality operator 47
P
Passato Prossimo 39
Passato Remoto 38
V
Varro 39