Mark Aronoff
Department of Linguistics
Stony Brook University
Stony Brook, NY 11794-4376
Email: mark.aronoff@stonybrook.edu
Telephone: 631-974-2394

1

2
Abstract
All linguists assume that the meaning of a complex syntactic expression is determined by its structure and the meanings of its constituents. Most believe that the meanings of words are similarly compositionally derived from the meanings of their constituent morphemes. The classical lexicalist hypothesis holds instead that the central basic meaningful constituents of language are not morphemes but lexemes. This article supports that hypothesis with evidence from syntax, lexical semantics, and morphology. Al-Sayyid Bedouin Sign Language (ABSL) is a new language that is compositional down to its smallest pieces, in which these pieces are lexemes. ABSL shows that a language can emerge very quickly in which lexemes are basic.
Next comes evidence against the claim that newly derived words diverge from compositionality only because they are stored in memory. Finally Hebrew verb roots are shown to have robust morphological properties that bear no relation to meaning and little to phonology, leaving room for morphemes within a lexeme-based framework, but not as the basic meaningful atoms of language.


This article is a revised version of my Presidential address, delivered at the annual meeting of the LSA in January 2007. Much of the content presented here is the result of joint work with
Frank Anshen, Robert Hoberman, Irit Meir, Carol Padden, Wendy Sandler, and Xu Zheng. I am grateful for discussion to my departmental colleagues at Stony Brook: Christina Bethin, Ellen
Broselow, Dan Finer, Alice Harris, Heejeong Ko, Richard Larson, and Lori Repetti; and to my former colleague Peter Ludlow. Thanks to the members of my immediate family for providing data and editorial assistance. This research was supported in part by the US-Israel Binational
Science Foundation, the National Institutes of Health, and the Stony Brook Provost’s Office.

3
Linguists have followed two ways in the study of words. One seeks to accommodate the word, the other to obliterate it. I defend here an approach to language that respects the autonomy of arbitrary individual words or lexemes and privileges the interaction between the idiosyncrasies of lexemes and a highly regular linguistic system.
i
Putting this approach in the context of the history of our field, I show that “Chaque mot a son histoire [Each word has its [own] history]”
(Gilliéron; see Malkiel 1964, p. 137) in “un système où tout se tient [a system where everything hangs together]” (Meillet 1903, p. 407), but instead of phonology, which was Gilliéron’s main concern, I concentrate on syntax, semantics, and morphology. I am not claiming that the lexeme is the only idiosyncratic element in language, but that it is the central such element.
1. Compositionality
All linguists assume that natural languages are compositional, which we understand informally to mean that the meaning of a complex syntactic expression is determined by its structure and the meanings of its constituents.
ii
The main argument for compositionality of natural language is rooted in syntactic productivity and was stated succinctly by Frege:
The possibility of our understanding sentences which we have never heard before rests evidently on this, that we can construct the sense of a sentence out of parts that correspond to words. (Frege 1914: 79)
Exactly what these “parts that correspond to words” are in natural languages has been debated since the beginning of linguistics (cf. the Ancient Greek grammatical term meros logou
‘piece/part of speech’). Saussure called them (simple) signs but he was not entirely clear about what linguistic units he took to be simple signs, though for the most part his examples are what

4 modern linguists would call lexemes (Saussure 1975; Carstairs-McCarthy 2005). A term like antidecompositionalist ‘opposed to the view that the meanings of words can be broken down into parts’, for example, can be analyzed into [[anti [[[de [com=pos-it]
V
]
V
ion]
N
al]
A
]
A
ist]
A
.
iii
The standard view is that the meanings of this word and others like it are compositionally derived from the meanings of these constituent morphemes, making morphemes, not words, the “parts that correspond to words” in natural languages. A bit confusing, but linguists like exoticizing the ordinary.
I am antidecompositionalist by temperament, because I believe that what happens inside lexemes
(Matthews 1972) is qualitatively different from what happens outside them. My adherence to this classic version of the LEXICALIST HYPOTHESIS (Chomsky 1970) is grounded in a simple love of words. As I wrote in my first publication (Aronoff 1972), on the semantics of the English word growth and its relevance to the lexicalist hypothesis, “all a person had to do [in order to notice the peculiar differences in meaning between the verb grow and the noun growth ] was to look the words up in a reputable dictionary, or think for a long time: grow – growth
” (p. 161).
Yet very few modern linguists either look words up in a reputable dictionary or think about them for a long time. Philosophers do, but linguists don’t often think much about the subtleties of word meanings, which is why most linguists are not antidecompositionalists.
The locus classicus of antidecompositionalist lexicalism is Chomsky 1970. This particular work, however, can be read in a myriad of ways, much in the way that most of Wittgenstein’s works can be read to support a great number of contradictory positions. What is beyond dispute is that
Chomsky introduced the terms lexicalist and lexicalist hypothesis in discussing derived nominals like refusal , eagerness , and confusion :

5
We might extend the base rules to accommodate the derived nominal directly (I will refer to this as the lexicalist position) … [T]here is no a priori insight into universal grammar … that bears on this question, which is a purely empirical one.
(Chomsky 1970, p. 188)
Chomsky brings three types of empirical evidence to bear on the issue of whether such derived nominals “are, in fact, transformationally related to the associated propositions”: restricted productivity, base-like syntax, and the fact that “the semantic relations between the associated proposition and the derived nominal are quite varied and idiosyncratic” (ibid. p. 188). He concludes that they are not transformationally related and that “The strongest and most interesting conclusion that follows from the lexicalist hypothesis is that derived nominals should have the form of base sentences” (ibid. p. 212), by which he means that derived nominals are inserted into lexical structures in the same way that simple base nouns are. This is the original lexicalist hypothesis, the idea that some members of major lexical categories (lexemes) are not derived by the same apparatus that derives sentences but are inserted into lexical categories in the base just as simple lexemes are.
iv
What I call the CLASSICAL LEXICALIST HYPOTHESIS
(Jackendoff 1975; Aronoff 1976) extends the scope of the original hypothesis to all lexemes and lexeme-internal sign structure (derivational morphology and compounding), though not to inflectional morphology (which relates forms of a single lexeme to one another), and claims that the level of the lexeme or lexical category is significant in the way that the level of the molecule is significant in chemistry: it has properties that are worth isolating. A NTIDECOMPOSITIONALISM is related to this classical lexicalist hypothesis: it says that the scope of syntax-based logical compositionality should not be extended below the lexeme because lexeme-internal structure
(however it is described) is different from lexeme-external structure.

6
An important and often forgotten point about antidecompositionalism is that, like Chomsky’s original lexicalist hypothesis, it is an empirical claim, not a theoretical postulate. Still, we antidecompositionalists do have theoretical obligations that our opponents do not, because we must account for the internal properties of lexemes in a different way from how we account for syntax (including inflection). That is what I have been trying to do for the last thirty-five years.
v
Also important is the fact that antidecompositionalists do not deny the value of morphemes, just as chemists do not deny the value of atoms and particles. There is plenty of evidence, linguistic and psycholinguistic, for morphemes and roots and for morphological relatedness. But none of this evidence, pace Marantz and Stockall 2006, supports a purely morpheme-based theory over one that recognizes lexemes but also recognizes roots and morphemes as morphologically significant elements, albeit not as reliable Saussurean signs.
vi
The rest of my article is divided into three parts. First, I discuss a language that appears to be completely compositional down to its smallest pieces, Al-Sayyid Bedouin Sign Language
(ABSL). I show that ABSL is an instantiation of Saussure’s picture of language, with little if any structure below the level of the lexeme. ABSL thus provides an unusual type of evidence for both compositionality and the lexicalist hypothesis, showing that very quickly, a language can emerge on its own in which words are indeed “the parts that correspond to words”.
I then turn to well-known phenomena from English and present evidence against one common misunderstanding of the special nature of words, that they diverge from compositionality only because they are stored in memory. I review old data and present one new example to show that even productively coined new words can be noncompositional.

7
I close with some traditional linguistic analysis and show that Semitic roots, the original lexemeinternal linguistic units, have robust properties that bear no relation to meaning and little to phonology. There is thus room for roots within a lexeme-based framework, but not as the basic meaningful atoms of language.
The approach to morphology that I take here is lexeme-based and realizational, using the taxonomy of morphological theories elaborated in Stump 2001. Like much morphological research of the last thirty years it is rooted in the work of P. H. Matthews (1972, 1974). It also has close affinities to Corbin 1984 and Anderson 1992. Carstairs-McCarthy 2002, Booij 2005, and Aronoff and Fudeman 2005 are recent introductions to morphology within this tradition.
Spencer 2000 provides a good brief overview. I do not think that it is unfair to say that the lexeme-based realizational approach has been prevalent among practicing morphologists for the last quarter century. Non-practitioners, however, seem to prefer morpheme-based approaches.
This raises the question of how much one should trust the experts, but I will not attempt to answer that question here.
2. Al-Sayyid Bedouin Sign Language
For several years, I have been privileged to be a member of a team, along with Wendy Sandler,
Irit Meir, and Carol Padden, working on Al-Sayyid Bedouin Sign Language (ABSL). ABSL has arisen in the last seventy years in an isolated endogamous community with a high incidence of nonsyndromic genetically recessive profound prelingual neurosensory deafness (Scott et al.
1995). What distinguishes ABSL from all other well-documented new languages are the circumstances of its creation and use, which show neither discontinuity of social structure nor the influence of other languages.

8
In the space of one generation from its inception, systematic word order has emerged in the language. This emergence cannot be attributed to influence from other languages, since the particular word orders that appear in ABSL differ from those found both in the ambient spoken languages in the community and in other sign languages in the area.
The Al-Sayyid Bedouin group was founded almost two hundred years ago in the Negev region of present-day Israel. The group is now in its seventh generation and contains about 3,500 members, all residing together in a single community exclusive of others. Consanguineous marriage has been the norm in the group since its third generation. Such marriage patterns are common in the area and lead to very strong group-internal bonds and group-external exclusion.
Within the past three generations, over 100 individuals with congenital deafness have been born into the community, all of them descendants of two of the founders’ five sons. Thus, the time at which the language originated and the number of generations through which it has passed can be pinpointed. All deaf individuals show profound prelingual neurosensory hearing loss at all frequencies, have an otherwise normal phenotype, and are of normal intelligence. Scott et al.
(1995) identify the deafness as (recessive) DFNB1 and show that it has a locus on chromosome
13q12 similar to the locus of several other forms of nonsyndromic deafness.
The deaf members of the community are fully integrated into its social structure and are not shunned or stigmatized (Kisch 2004). Both male and female deaf members of the community marry, always to hearing individuals. The deaf members of the community and a significant fraction of its hearing members communicate by means of a sign language. Siblings and children of deaf individuals, and other members of a household (which may include several wives and their children) often become fluent signers.

9
Members of the community generally recognize the sign language as a second language of the village. Hearing people in the village routinely assess their own proficiency, praising those with greater facility in the language. Those who have any familiarity with Israeli Sign Language, including those who have attended schools for the deaf outside the village, recognize that the two sign languages are distinct. Nor do Al-Sayyid signers readily understand the Jordanian sign language used in simultaneous interpreting on Jordanian television programs received in the area.
Many of the signers in this community are hearing, a highly unusual linguistic situation, but one that is predicted to arise as a consequence of recessive deafness in a closed community (Lane,
Piller and French 2000). One result of the recessiveness is that there are a proportionately large number of deaf individuals distributed throughout the community (over 4 percent). This means that more hearing members of the community have daily contact with deaf members, and consequently signing is not restricted to deaf people. Furthermore, each new generation of signers is born into a native-like environment with numerous adult models of the language available to them. ABSL thus presents a unique opportunity to study a new language that has grown inside a stable community without obvious external influence. We have identified three generations of signers. The first generation in which deafness appeared in the community (the fifth since the founding of the community) included fewer than ten deaf individuals, all of whom are deceased. Information on their language is limited to reports that they did sign and one very short videotape record of one of these individuals. I report here only on the language of the second generation.
I discuss the order of signs within sentences and within phrases in the language, based on a tally of all sentences in a database that consist of more than one sign (Sandler et al. 2005). Sentences

10 consisting solely of a predicate and utterances consisting solely of nouns were set aside. We first tallied the order among the major elements of the remaining clauses, the predicate (henceforth V) and its arguments. Table 1 shows the order of predicates relative to arguments within the clauses considered in our count. Out of 158 clauses, 136 are predicate final.
INSERT TABLE 1 ABOUT HERE
Table 2 shows the frequency of the relative orders of S, O, IO, and V within the data (following standard practice, we lump together O and IO unless a clause contains both). One hundred twenty six clauses contain only one noun, of which 58 contain S and V, and 68 O and V. Where two nouns are mentioned in a clause, S precedes O (in all 32 cases). S never follows V. Thus we conclude that the prevalent word order for ABSL sentences is SOV. A binomial test yields statistically reliable p-values of less than .00001 for this and all other word order effects reported here.
INSERT TABLE 2 ABOUT HERE
We next considered the order of modifier elements within phrases (adjectives, negatives, and numerals) relative to their head nouns and verbs. As shown in Table 3, in all instances but one, the modifier follows its head.
INSERT TABLE 3 ABOUT HERE
Overall, we find structural regularities in the order of signs in the language: SOV order within sentences, and Head-Modifier order within phrases.
vii
These word orders cannot be attributed to the ambient spoken language. The basic word order in the spoken Arabic dialect of the hearing members of the community, as well as in Hebrew, is SVO. This generation of signers had little or no contact with Israeli Sign Language, whose word order appears to vary more widely in any

11 case. Nor can the Head-Modifier order be ascribed to the ambient colloquial Arabic dialect of the community. In this dialect, and in Semitic languages generally, although adjectives do follow nouns, numerals precede nouns; and negative markers only rarely follow their heads.
Hence the robust word-order pattern exhibited by the data is all the more striking, since it cannot be attributed to the influence of other languages; rather, this pattern should be regarded as an independent development within the language.
This remarkable structural clarity breaks down entirely when we look inside words, where we find very little structure, either morphological or phonological. The language does have a fair number of compounds, but compounds are not morphological in the strict sense, since they are composed entirely of words and do not contain any bound elements.
viii
Sign languages are known for their robust inflectional morphology, especially their complex verb agreement systems, which arise very quickly in the history of individual languages (Aronoff, Meir, and
Sandler 2005). Our team was therefore surprised to discover that ABSL has no agreement morphology, indeed no inflectional morphology at all (Aronoff, Meir, Padden, and Sandler
2005). Nor does it appear to have a phonological system of the kind we are familiar with in other sign languages.
Phonology is an instantiation of what Hockett (1960) calls duality of patterning, or what others have called double articulation (Martinet 1957). When we say that a sign language has phonology, we are saying that it has a set of discrete meaningless contrastive elements
(handshapes, locations, and movements) that form their own system, with constraints on their combination, constituting a second articulation independent of the first articulation of meaningful elements. This system of contrastive elements makes up the signifiant of the Saussurean sign. It

12 was the existence of such a system in American Sign Language that first led William Stokoe to proclaim its legitimacy as a language (Stokoe 1960).
ABSL appears not to have a system of discrete meaningless elements within words. Instead, each word has a conventionalized form, with tokens roughly organized around a prototype, but no internal structure, as the following illustrations of the ABSL signs KETTLE (figure 1) and
BANANA (figure 2) illustrate. KETTLE is a compound, of which the first component is CUP and the second either POUR or BOIL. For lack of space, it is impossible to show more than a few examples of each, but in both, and especially in BANANA, there is considerable variation among individual signers. The variation found among signers in many words in our data (of which BANANA is a good example) lies precisely along the parameters that create contrasts in other sign languages. The difference in the amount of variance between KETTLE and
BANANA may also be taken as an indication that certain signs, like KETTLE, are becoming more conventionalized, but conventionalization does not equal phonology. It is also noteworthy that the signers in figures 5c-e, whose sign for BANANA is very similar, are brothers. We do find much more similarity in sign form within single families.
INSERT FIGURES 1 AND 2
ABSL has been able to develop into a full-fledged linguistic system without benefit of phonology because of the visual medium of signing, which has many more dimensions than sound does and which allows for direct iconicity (Aronoff, Meir, and Sandler 2005). As Hockett notes (1960, p. 95):
There is excellent reason to believe that duality of patterning was the last property to be developed, because one can find little if any reason why a communicative system should have this property unless it is highly complicated. If a vocal-auditory system (emphasis

13 added/MA) comes to have a larger and larger number of distinct meaningful elements, those elements inevitably come to be more and more similar to one another in sound.
There is a practical limit, for any species or any machine, to the number of distinct stimuli that can be discriminated, especially when the discriminations typically have to be made in noisy conditions.
It may be that ABSL will develop phonology as it ages, perhaps simply as a function of the size of its vocabulary, as Hockett suggests (see Nowak and Krakauer 1999 for a mathematical model). Indeed, our team is currently investigating questions of just this sort. For the moment, though, ABSL is a language with a compositional syntax, but very little word-internal structure of any kind. It is, from a certain perspective, a perfect language (Chomsky 1995, 2000; Eco
1995).
3. The Meanings of morphologically complex words
The tenet that every natural language is a unitary compositional system of sentences whose basic building blocks are morphemes has an intellectual sister in the notion that the meaning and structure of morphologically complex words can not merely be represented in terms of their sentential paraphrases but should be formally reduced to these paraphrases, so that the words themselves are reduced to sentences or propositions and morphology to syntax. To proponents of this method, words and sentences really are the same kinds of things, because words are sentences in miniature, linguistic munchkins. This idea dates back at least to Panini and it has been popular among generativists almost from the beginning; not a surprise, since the most successful early generative analyses, notably those of Chomsky 1957, make no distinction between syntax and morphology, at least inflectional morphology.
ix

14
Precisely those morphologically complex word types that have seemed most akin to sentences in the structure of their meanings, however, when studied in more detail, provide empirical support for just the opposite conclusion: morphologically complex words are not sentences and their meanings are arrived at in an entirely different fashion. Almost all the evidence that I review is over twenty-five years old and much of the field has successfully ignored it, so I am very much in the minority. Nonetheless, I would like to take advantage of the occasion to make a new plea for what, to my mind at least, have been very important findings.
The founding generative work in the enterprise of reducing words to sentences, as Chomsky notes (1970, p. 188) is Lees 1960.
x
The details of Lees’s analysis have been forgotten, but his idea has flourished on and off ever since, more on than off, largely unchanged except in notation.
I offer two examples of Lees’s analysis. The first is what he calls action nominals. Here are two sample such nouns and their sentential sources, according to Lees:
1) The committee appoints John

John’s appointment by the committee
2) The committee objects to John

The committee’s objection to John
These and others with different morphology are produced by means of a single action nominal transformation. The second example is nominal compounds. Here Lees discusses eight types, which include subject-predicate, subject-verb, verb-object, and others, all based on syntactic relations. His discussion of these types takes up fifty pages, so I can’t paraphrase it here, but a few samples are given below:
3) The plant assembles autos

auto assembly plant
4) The cap is white

white cap
5) The artist has a model

artist’s model

15
6) The sheep has a horn. The horn is like a prong

pronghorn
Linguists are tempted to try to reduce the structures and meanings of complex words to those of sentences because they have a good set of tools for studying the syntax and semantics of sentences, which they apply to words. But words have much less apparent internal structure.
This difference is both a blessing and a curse: we can transfer with ease what we know about the internal structure and meanings of sentences to words, but we have no direct way of finding out the actual extent of the isomorphism between the two domains. We may very well be deluding ourselves.
A second problem is that, unlike sentences, words are not essentially ephemeral objects. As
Paul Bloom has shown, though syntactic development stops fairly early, certainly before puberty, individual speakers of a language continue to accumulate words at an astonishing rate throughout their lives and retain them in memory (Bloom 2000). They do not accumulate sentences. Because sentences are ephemeral, their meanings must be compositional sums of the meanings of their parts. This was Frege’s main argument for the compositionality of sentences in natural languages. By the same token, because words are not so ephemeral, and more often than not retained in memory and society, complex words do not have to be compositional. Words that have been around for any time at all develop idiosyncrasies that are passed on to their new learners.
There have been two lines of response to this obvious noncompositionality of some complex words. The more common response is to say that complex words are compositional at heart and that the departures from regularity accrete precisely because words are stored and used. The semantic differences between sentences and complex words are thus purely accidental on this view, the result of nonlinguistic or noncomputational factors.

16
One example in support of the position that words aren’t born with noncompositional meaning but have it thrust upon them is the trio of words cowboy , refugee , and skinner , which are curiously synonymous in one sense: ‘one of a band of loyalist guerillas and irregular cavalry that operated mostly in Westchester County, New York, during the American Revolution’ ( Webster’s
Third New International Dictionary , henceforth WIII). Explaining why these three words came to be used to denote this particular group is a philological adventure of precisely the sort that gives linguists a bad taste for etymology, but it is obvious that that story has nothing to do with the compositional meaning of any of them. Instead, this peculiar meaning arose for all three words under certain very particular historical circumstances. The fact that this meaning of these words is not compositional is, linguistically speaking, an accident. The meaning itself is no more than a curiosity. This analysis allows us to preserve the intuition that words are compositional at birth, just like sentences, but that they lose their compositional meanings and acquire idiosyncratic meanings like this one just because they are stored and used.
I followed this well-trodden road at first, firm in the faith that the meanings of productivelyformed words would turn out to be compositional if we could ever catch the words at their moment of birth. But I found that road too easy and so I took the one less traveled by, and that has made all the difference. This less traveled road follows a Miesian method, taking the surface structure of complex words at face value and replacing the analogy to sentences with a very sparse non-sentential syntax that mirrors this simple structure. The semantics of this sparse syntax for complex words is compositional, but only because it accounts for much less of the actual meaning of even newborn words than other semantic systems attempt to cover, leaving the rest to pragmatics (Aronoff 1980). This bifurcating treatment rests on the assumption that words have complex meanings precisely because neither words nor their meanings are entirely

17 linguistic objects, but rather the bastard offspring of language and the real or imagined world; it is this union of sparse linguistic resources with the vastness of the nonlinguistic universe that makes all words so rich from birth. Noncompositionality, on this account, follows from the nature of words, not from their nurture. Of course, new complex words cannot be entirely arbitrary in meaning, because, except for Humpty Dumpy, we use them expecting our interlocutors to understand us.
xi
Some are more predictable than others, but new complex words are never entirely the product of language.
As James McClelland has kindly pointed out to me, on this view the meanings of words may be compositional just in case the world that these words and their meanings intersects with is itself a logical world. A good example is chemical terminology. It is well known that this terminological system is compositional, but this follows directly from the fact that the system is itself entirely determined by the scientific principles of chemistry. Indeed, the name of any chemical compound is governed by the principles of the Interdivisional Committee on
Terminology, Nomenclature and Symbols (ICTNS) of the International Union of Pure and
Applied Chemistry. Not surprisingly, this terminology can yield very long words, the longest one attested in print in English being acetylseryltyrosylserylisoleucylthreonylserylprolylserylglutaminylphenylalanylvalylphenylalanylleucylserylserylvalyltryptophylalanylaspartylprolylisoleucylglutamylleucylleucylasparaginylvalylcysteinylthreonylserylserylleucylglycylasparaginylglutaminylphenylalanylglutaminylthreonylglutaminylglutaminylalanylarginylthreonylthreonylglutaminylvalylglutaminylglutaminylphenylalanylserylglutaminylvalyltryptophyllysylprolylphenylalanylprolylglutaminylserylthreonylvalylarginylphenylalanylprolylglycylaspartylvalyltyrosyllysylvalyltyrosylarginyltyrosylasparaginylalanylvalylleucyl-

18 aspartylprolylleucylisoleucylthreonylalanylleucylleucylglycylthreonylphenylalanylaspartylthreonylarginylasparaginylarginylisoleucylisoleucylglutamylvalylglutamylasparaginylglutaminylglutaminylserylprolylthreonylthreonylalanylglutamylthreonylleucylaspartylalanylthreonylarginylarginylvalylaspartylaspartylalanylthreonylvalylalanylisoleucylarginylserylalanylasparaginylisoleucylasparaginylleucylvalylasparaginylglutamylleucylvalylarginylglycylthreonylglycylleucyltyrosylasparaginylglutaminylasparaginylthreonylphenylalanylglutamylserylmethionylserylglycylleucylvalyltryptophylthreonylserylalanylprolylalanylserine (a strain of tobacco mosaic virus). My point is that chemical terminology is compositional, not for linguistic reasons, but because the terminology mirrors chemical theory, which is a logical world. But when the world that the words reflect is not entirely logical (and few are), then neither are the meanings of the words.
To show what motivates this bifurcating treatment of the meanings of words, I return to one of our trio of Revolutionary words, skinner , the one that appears to have traveled furthest from its compositional roots to arrive at this peculiar meaning. I do not attempt to elucidate the meaning under consideration, whose origin is entirely opaque, but look only at the meaning that seems to be most rooted in language and least tied to language-external context. It quickly becomes apparent that even that simple attempt comes to grief.
A skinner is transparently ‘one who skins’ and it is easy to derive the most common meaning of the noun skinner from that of the verb skin via just about any syntactic or semantic theory. So far, so good, but move one step further back to the meaning of the verb and the going gets tough.
This verb skin is derived from the noun skin , whose Old Norse origin tells us it has been around for over a millennium. How are the two related semantically? Here are a few of the 26 senses of

the verb listed in
Webster’s Third New International Dictionary
. I have omitted all those that are apparent extensions of others:
7)
19 a) To cover with or as if with skin <fuselages and wings will be skinned with steel . . . or titanium alloys –Werner von Braun> b) To heal over with skin c) To strip, scrape, or rub off the skin, peel, rind, or other outer coverings of: to remove a surface layer from <huge catfish are skinned and dressed by hand. . . > d) To remove (skin or other outer covering) from an object: pull or strip off <too late to skin out the hide that night – Corey Ford> e) To chip, cut, or damage the surface of < skinned his hand on the rough rock> f) To outdistance or defeat in a race or contest g) To equalize the thickness of adhesive on (a pasted or glued surface) by placing a sheet of wastepaper over it and rapidly rubbing or pressing h)
To become covered with or as if by skin <these inks won’t dry in the press . . . nor will they skin in the can> – usually used with over . i) To climb or descend – used with up or down < skinned down inside ladders from the bridge deck –K. M. Dodgson> j) To pass with scant room to spare: traverse a narrow opening – used with through or by
<the big ship barely skinned through the open draw>

20
Rest assured I did not choose this word because of its many senses. Truthfully, I did not choose it at all, and the whole excursion that led to the curious incident of the Westchester synonyms began quite by accident, during the course of a conversation about a group the country later chose to call hurricane Katrina evacuees, on the grounds that the term first used, refugee , was pejorative. Such rich variety of senses is simply typical of zero-derived denominal verbs in
English (henceforth zero-verbs). Clark and Clark (1979) provide a catalog of the major categories of sense types for zero-verbs, based on their paraphrases, which number in the dozens.
A synopsis is given in Table 4.
INSERT TABLE 4 ABOUT HERE
In the face of such a wealth of sense types, the syntactically inclined explainer has no choice but to deny all the best Ockhamite tendencies and resort to a multiplicity of tree-types, each with the noun or root in a different configuration. Lees would have done so, as he did for compounds, and as have more recent scholars, notably Hale and Keyser (1993) and their followers.
xii
Even if we ignore the problem of unrecoverable deletion that plagues these analyses, the proliferation of syntactic and semantic types assembled under a single morphological construction is embarrassing.
But how can one unite all these types of senses? Jespersen is prescient, as he often is. In the volume on morphology of A Modern English Grammar on Historical Principles (vol. 6, 1943), he has the following to say about the meanings of compounds:
Compounds express a relation between two objects or notions, but say nothing of the way in which the relation is to be understood. That must be inferred from the context or otherwise.
Theoretically, this leaves room for a large number of different interpretations of one and the

21 same compound . . . The analysis of the possible sense-relations can never be exhaustive.
(137-38)
The only method that has been used successfully to simultaneously avoid and explain this choice among and infinity of senses removes both the complexity and the variance from the linguistic to the pragmatic realm, as Jespersen suggests and as I am advocating. Pamela Downing pioneered it in her 1977 article on English nominal compounds. As Downing puts it, echoing Jespersen:
“Indeed, because of the important differences in the functions served by compounds, as opposed to the sentence structures which more or less accurately paraphrase them, attempts to characterize compounds as derived from a limited set of such structures can only be considered misguided. A paraphrase relationship need not imply a derivational one.” (840-841) She concludes that there are “no constraints on the N + N compounding process itself.” (841). The constraints lie instead in how people categorize and refer.
English zero-verbs show a similarly rich and varied set of interpretations, as we have seen in the case of skin
. Eve and Herb Clark come close to an account like Downing’s for these verbs, but they confine themselves to innovations, which they call CONTEXTUALS . In my response to the
Clarks (Aronoff 1980), I suggested that it is possible to account uniformly for all zero-verbs, not just the innovations, by a conversion (rule) of the simple form N

V and that the meaning of the innovative verb always comprises what I call an evaluative domain of the noun’s denotation
(essentially a dimension along which the denotation of the noun can be evaluated; a knife is good if it cuts well, a mother is good if she does well what mothers do; a club is good for clubbing, etc.) For most nouns or other lexical items, there is no fixed evaluative domain, so that what the meaning of the novel zero-verb will be depends on the context of its use. Of course, any word’s meaning will become fixed lexically with enough use and time, but that fixing should be of no

22 interest to a linguist. This story holds most remarkably for verbs like boycott and lynch , which are derived from proper nouns and whose meanings are traceable to very specific incidents in which the named person, here Boycott or Lynch, played an important role. There is nothing else to say about the semantics of zero-verbs. Even the notion of an evaluative domain is superfluous, since Gricean principles dictate that the verb have something to do with the noun, no more and no less than what needs to be said to account for the range of data. On this analysis, then, all that the grammar of English contains is the noun-to-verb conversion.
The beauty of this kind of account is that it leaves the words largely untouched, freeing them to vary as much as speakers need or want them to. Critics have replied that this variance is no beauty and that people like me who advocate such a sparse account, few though we may be, are simply irresponsible and should be purged from the field. These critics, though, ignore the must important point about Downing’s conclusion and mine, which is that they are not theoretical but empirical. When we look at the actual meanings that speakers attribute to novel compound nouns under experimental conditions, as Downing did, we find that they vary in ways that are formless and void. “The constraints on N + N compounds in English cannot be characterized in terms of absolute limitations in the semantic or syntactic structures from which they are derived”
(840). Words have idiosyncratic meanings not just because they are preserved in memory and society but because words categorize and refer outside language.
I offer one more English zero-verb to buttress my point. The verb is friend , which has been cited on and off for at least a century, but has never caught on much until very recently. The one definition given in WIII is ‘to act as the friend of’ and WIII cites a line from
A Shropshire Lad
(1896): <and I will friend you, if I may, in the dark and cloudy day – A. E. Housman>. In the last couple of years, though, friend has reemerged as a common term among the myriads who

23 use the websites friendster.com and facebook.com, both of which crucially involve individual members maintaining lists of ‘friends’. Though the term seems to have originated with the earlier Friendster, I concentrate on Facebook, which began, by its own definition, as ‘an online directory that connects people through social networks at school’, because I have the best ethnographic sources of data for it (three children of the right age and their friends). Facebook was inaugurated only in late 2003, but the community of members is quite large and active and has been the subject of articles in major media outlets. To friend someone within the Facebook community is not to act as a friend to that person, but to invite someone to be your Facebook friend, which you do by trying to add that person to your list of Facebook friends. Here is how that works, according to the Facebook help page (from late 2005): xiii
You can invite anyone that you can see on the network to be your friend. Just use the "Search" page to find people you know and then click on the "Add to Friends" button on the right side of the screen. A friend request will be sent to that person. Once they confirm that they actually are friends with you, they will show up in your friends list.
Here are some uses of friend in this sense: xiv
8) a) EWW! This guy from my London seminar who's a total ASS just friended me! b) All these random people from high school have been friending me this week! c) Should I friend that dude we were talking to at 1020 last night? I don't want him to think
I'm a stalker.
The invitation is crucial to the meaning of friend v
, because the reciprocity of the friend relation in
Facebook is enforced by the system: your invitee is not added to your list of friends unless they

24 accept the invitation. Oddly, if the invitee does not accept, “the [inviter] will not be notified.
They also will not be able to send you [the invitee] another friend request for some amount of time, so to them, it will just seem as if you haven't confirmed their friendship yet.” The creators of Facebook may see this as a polite method of rejection. I find it odd, but I am not twenty years old, so I guess I just don’t understand.
The point of all this is that the meaning of friend x can not just be ‘act as a friend of x’, as it clearly is in the Housman citation (note the use of the modal may there) but the invitation, which we may think of as conative aspect, must be built into the meaning. To friend x in this world is
‘to try to become a friend of x’ or ‘to ask x to be one’s friend’, in the special sense of the noun friend that applies in this world. Of course, we know why the extra predicate is part of the meaning of the verb: it is built into the program and therefore into the social community that surrounds the program! One could try to amend the Clarks’ list or its syntactic equivalent to allow for multi-predicate or aspectual meanings, but that would avoid the underlying problem, that the meaning depends on extralinguistic factors from the very first moment of the coining of the word, which is precisely Downing’s empirical point.
xv
4. Lexical roots
The lexeme-centric view of word meaning, where lexical semantic information is at least partly nonlinguistic and does not reside in morphemes, has implications for the meanings of roots. If I am right about how lexical semantics works, then two words can have the same root and not share much lexical meaning. There is little relation in meaning, for example, between the name
Bork and the verb bork , yet they are as closely related morphologically as any two words can be and must share a root, if the term root is to have any content. What does one say then about

25 lexical roots, which are supposed to be the atomic meaningful units of language, if two instances of the same root can share so little meaning? The simplest ploy is to deny the linguistic reality of roots entirely, what my colleague Robert Hoberman calls the anti-rootarian position. But there is a middle ground, where words have morphological structure even when they are not compositionally derived, and where roots are morphologically important entities, though not particularly characterized by lexical meaning.
4.1. Latin verb roots
In the first piece of morphological analysis I ever did (published in Aronoff 1976), I showed that certain Latin-derived roots in English verbs can be active morphologically even when they are obviously meaningless: each root conditions a certain set of affixes and alternations that is peculiar to that root alone. Some examples are given in Table 5.
INSERT TABLE 5 ABOUT HERE
These Latinate roots in English are the historical reflexes, through a complex borrowing process, of the corresponding verb roots in Latin. Even for Classical Latin, the alternations exhibited by at least some of these verb roots were not phonologically motivated. But in Latin, these same roots were morphologically active in other ways too that are independent of syntax or semantics.
Perhaps most intriguing is the role of roots in determining whether a given verb was
DEPONENT
.
The traditional Latin grammatical term deponent is the present participle of the Latin verb deponere ‘set aside’. Deponent verbs have set aside their normal active forms and instead use the corresponding passive forms (except for the present participle) in active syntactic contexts.
The verb a dmetior
‘measure out’, for example, is transitive and takes an accusative object, but passive in form, because it is deponent. Similarly for obliviscor ‘forget’ (oblitus sum omnia ‘I

26 have forgotten everything’ [Plautus]), scrutor
‘examine’, and several hundred other Latin verbs.
There are even semi-deponent verbs, which are deponent only in forms based on the perfect stem.
For some time, Xu Zheng, Frank Anshen, and I have been working on a comprehensive study of
Latin deponent verbs, using a database of all the main-entry deponent verbs in the Oxford Latin
Dictionary , which covers the period from the first Latin writings through the end of the second century CE (Xu, Aronoff, and Anshen 2007). I report here on only one small part of our research, that involving deponent roots.
We show in Xu et al. 2007, based on an exhaustive analysis of all the senses of all deponent verbs, that neither syntax nor semantics is the best predictor of deponency, though both are factors. Our database contains 287 deponent verbs not derived from another lexical category
(about half the total number of deponent verbs), the great majority of them consisting of either a bare root or a root and a single prefix. The senses of all deponent verbs are about evenly divided between transitive and intransitive values and a given root may appear in both transitive and intransitive verbs. Of these 287 verbs not derived from another lexical category, 85% (244) have deponent roots, which we define as a verb root that occurs only in deponent verbs. There are 52 deponent roots, 22 of which occur in four or more distinct verbs. Table 6 is a list of these roots.
Thus, whether a given verb will be deponent is determined to a great extent by its root, independent of the meaning of either. In other words, besides the alternations that have survived in English and the Romance languages, deponent Latin verb roots have at least one other important morphological property (their mere deponency) that is independent of either syntax or semantics.
INSERT TABLE 6 ABOUT HERE

27
4.2. Hebrew Roots
I now turn to the role of the root in the morphology of Modern Hebrew xvi
verbs. For more than a millennium, traditional grammarians have characterized Semitic languages as being morphologically grounded in meaningful ROOTS , each root consisting usually of three consonants, with some number of lexemes (verbs, nouns, and adjectives) built on each root.
But not every Modern Hebrew root has a constant meaning.
xvii
Tables 7 and 8 each contain a single root with an apparently constant meaning.
xviii
The meanings of the individual noun, verb, and adjective in these tables can be predicted reasonably well if each morphological pattern is associated with its own compositional semantic function. This is what we expect from the traditional account. Table 9, though, quickly brings such an enterprise to grief. What do pickles
( kvu∫im
) have to do with highways ( kvi∫im
)? The story of how these two words are related is quite simple. Paved roads in early modern Palestine were macadamized: made of layers of compacted broken stone bound together with tar or asphalt. Modern roads are still built by pressing layers, but in a more sophisticated manner. Traditional pickling also involves pressing: whatever is to be pickled is immersed in brine and pressed down with a weight, but the container should not be sealed. If it is, it may explode (as I know from personal experience). But pressing alone without brine does not constitute pickling. Every cook worth his or her salt knows that beef brisket should be pressed after it is cooked and before it is sliced, regardless of whether it is pickled (corned) or not. Modern industrial pickling does away with pressing in various ways and not every paved road in Israel is called a kvi∫ , only a highway is, so while it is clear why pickles and highways both originate in pressing, there is nothing left of pressing in the meanings of these Hebrew words today. One who points out their related etymologies to a native speaker

28 of the language is rewarded with the tolerant smile reserved for pedants, as I also know from personal experience.
INSERT TABLES 7 – 9 ABOUT HERE
The reflex response to problems like this is to posit an ‘underspecified’ core meaning for the root, which is supplemented idiosyncratically in each lexical entry. It is logically impossible to show that underspecification is wrong, but trying to find a common meaning shared by pickles and highways brings one close to empirical emptiness and this methodological danger recurs frequently in any Semitic language. In any case, there is no need to find a common meaning in order to relate the two words morphologically, as I show. In fact, Hebrew verb roots can be identified on the basis of alternation classes remarkably similar in type to those that operate on
Latinate roots in English, without reference to meaning or regular phonology. This observation is nothing new; these alternation classes are the bane of Hebrew students.
xix
First, we must know a few basics of Hebrew verb morphology. Hebrew has a set of what are traditionally called binyanim , seven in number, often called conjugations in the recent theoretical literature. Each binyan rigidly assigns to each cell of the verb paradigm a stem pattern containing a prosodic shape complete with vowels. A stem pattern may also include a prefix.
The hif`il binyan, for example, has a prefix hi - or ha -, and the pattern CCiC (e.g. higdil ‘grew’).
The pi`el binyan has the stem pattern CiCeC in the past, meCaCeC in the present, and CaCeC in the future (e.g. megadel ‘grow’). Traditionally, each binyan (like each root) is said to have a constant meaning or syntactic structure. Some modern scholars have questioned that claim, but whether binyanim have meaning is orthogonal to the matter at hand, which is roots.
Besides being determined by binyanim, verb forms depend on the alternation class of their root.
Root alternation classes, except for the (default) regular class, are traditionally identified in terms

29 of one of the consonants that occupy the three canonical root consonant positions, which I call
R
1
, R
2
, and R
3
.
xx The term
FULL
is used for regular roots because they do not alternate except for more general phonological phenomena that are not root-dependent, like spirantization (the alternation of stops and fricatives). For example, the root alternation class R
3 h is so named because the third consonant of the root is h ; in a root of alternation class R
3 n the third consonant is n . These alternation classes, named in this way, figure prominently in traditional Hebrew grammar.
At some point in the early history of the language, the differences in verb forms among root alternation classes were undoubtedly predictable from the phonology of the consonant that defined each type. R
1 n verbs, which we look at in some detail below, pattern the way they do because of the phonology of n in some very early stage of Hebrew (long before the original consonantal text of the Bible or any other existing Hebrew text was written down). The passage of time, however, made these phonological conditions opaque, so that two verbs with identical consonants in the same position could show different alternation patterns even in Biblical
Hebrew. This differential patterning proves that the alternation classes were no longer predictable phonologically, even in Biblical Hebrew, but are more like verb classes of the sort found, for instance, in Germanic languages, where the classes are defined by distinctions in ablaut and verbs must be marked for membership in a given STRONG class.
xxi
In Hebrew similar sorts of idiosyncratic alternations to those of Germanic are determined by the alternation class of a verb’s root (which also appears to be true in Germanic), a fact that gives roots their linguistic reality.
The root alternation classes and binyanim of Semitic constitute two dimensions of a fourdimensional matrix, the others of which are person/number/gender combination (ten are

30 possible), and tense (in Modern Hebrew: past, present, future, and imperative, though the last is becoming increasingly rare). All the possible regularities of any given individual verb form are thus exhausted by four specifications: its root class, its binyan, its person/number/gender value, and its tense. Lest it be thought that this matrix is trivial, a reasonably complete traditional table of verbs (Tarmon and Uval 1991) lists 235 distinct combinations of root class and binyan alone, what they call verb types, of which only 9 contain just a single root. Multiplying 235 by the ten person/number/gender values in all of the tenses (a total of 26) yields a little more than six thousand distinct form types into which any verb form in the language must fall.
xxii
Not all of these six thousand types must be learned individually, only what alternation class a verb root belongs to (if it is not regular) and what binyanim a root may occur in.
In this article, I look only at the roots whose R
1
is a coronal sonorant: R
1 n , R
1 y , and R
1 l . My task is to identify the marked alternation classes that such roots belong to, those alternation classes whose member roots must be listed (weeding out, along the way, those coronal sonorant-initial roots that do not belong to any marked class and are hence regular). Only the class of full roots is unmarked. It constitutes the default class for Hebrew roots. And just like default classes in other languages, it includes those roots that exceptionally do not pattern with a marked class, although they meet the criteria for membership in that class (Aronoff 1994). For example, there are two morphologically distinct sets of R
1 n roots and they cannot be distinguished by their phonological makeup, which means that the R
1 n roots in one set must be marked as belonging to a marked alternation class. Because one set patterns exactly like regular full verbs, the other set constitutes the marked alternation class. Any root that belongs to a marked class alternation must be flagged as such, in precisely the same way that a verb is flagged for its conjugation in a classical Indo-European language.

31
To a great extent, root alternation classes resemble the conjugation classes of classical Indo-
European languages more than binyanim do, even though theoretical linguists usually call the latter conjugations. One might say that there are two dimensions to the conjugation of a Semitic verb, the root class and the binyan. This is certainly close to the traditional view of Hebrew grammar.
To demonstrate the validity of this claim, I show first that there are two sets of R
1 n roots and that one set behaves just like regular roots, making the other set a marked class of roots. I then show that there are three sets of R
1 y roots, one of which patterns identically to the marked R
1 n alternation class, leading to the conclusion that the so-called R
1 n alternation class contains verbs with both initial n and y , (and one with initial l ), not news to traditional grammarians.
In other words, the alternation class of roots that includes marked R
1 n roots also includes roots whose R
1 is not n . This alternation class has thus become morphological rather than phonological. One might wish to give the alternation class formerly known as R
1 n a more general label: the class of marked R
1 coronal sonorant roots, but the existence of a distinct class of R
1 y roots makes that label unenlightening. The main point is that the class has become morphological, even by the earliest attestation of Biblical Hebrew.
The two morphologically distinct sets of R
1 n roots are those in which the n is deleted between the vowel of a prefix and R
2
, which are traditionally termed missing R
1 n roots, and those in which the n is not deleted. The latter pattern exactly as do completely regular roots, and thus are actually full roots, not members of a marked alternation class, exactly as predicted generally for exceptions to marked classes (Aronoff 1994). Originally, the difference in patterning was phonologically motivated: an n went missing because it was assimilated to a following nonguttural R
2
consonant, which was thereby geminated and, if a stop, consequently exempt from

32 spirantization (realization as a fricative).
xxiii
The roots npl , for example, when prefixed with the first person plural imperfect prefix ni -, would lose its initial n , so that the actual Masoretic form was nippol
‘we (will) fall’. xxiv
In roots with guttural (pharyngeal or laryngeal) R
2
, the n was not assimilated, being somehow protected by the guttural. The root nhg , for example, when prefixed with the first person plural imperfect prefix ni -, would not lose its initial n , so that the actual
Masoretic and Modern form is ninhag
‘we (will) drive’. The assimilation of n has been unproductive in recorded history, however, so that, in a good number of roots with a non-guttural
R
2
, the R
1 n remains intact even in the assimilation context. Whether R
1 n is missing is a property of individual verb roots, not lexemes: npl is a missing R
1 n root; ngd is not.
xxv
This is the principal empirical analytical evidence for the linguistic reality of Hebrew roots. We even find etymologically distinct homophonous roots, one of which is a missing R
1 n root, and the other not. Compare hibit ‘look’ with hinbit ‘sprout’.
xxvi Importantly, individual lexemes belonging to the same root do not vary. Examples of missing R
1 n roots are given in Table 10; full R
1 n roots are exemplified in Table 11. Individual verb forms in which the n is missing are italicized; those in which it is phonologically eligible to be missing but isn’t are boldfaced.
INSERT TABLES 10 - 11 ABOUT HERE
There are three sets of R
1 y roots, distinguished by their hif`il binyan forms: i.
In the most numerous set, the vowel of the hif’il prefix is o instead of i throughout the hif`il paradigm. This prefix vowel occurs only with this root alternation class and only in the hif`il binyan). Also, the y is missing. Because it is most numerous, I call this the missing R
1 y alternation class.

33 ii.
In three roots, the hif`il paradigm is identical to the missing R
1 n root class paradigm throughout.
xxvii These are clearly members of the missing R
1 n root class, even though their R
1
is y and not n . iii.
In four roots, the y is never deleted and the hif’il prefix vowel is e .
Table 12 compares the three varieties of R
1 y roots with R
1 n roots.
INSERT TABLE 12 ABOUT HERE
The four R
1 y roots in which the y never deletes in hif’il are best viewed as full roots, parallel to full R
1 n roots (with the prefix vowel additionally becoming e for phonological reasons). One of these roots ( ynq ) and a few others appear in the pa`al binyan with the y intact and show the normal i vocalism of full roots in pa`al (e.g. yiynaq , yiyra’
, yiy∫an ), lending further support to their classification as full roots.
xxviii
The missing R
1 y roots that follow the local default o vocalism in hif`il and also occur in the pa`al binyan show two traits in their pa`al that further characterize this missing R
1 y alternation class: unlike all other root alternation classes, they have an e: vocalism throughout the future and the y drops in both the future and the imperative.
To summarize, I have shown that there are two marked alternation classes of coronal-sonorantinitial verb roots in Hebrew and that, though these are traditionally labeled R
1 n and R
1 y , that characterization is not entirely correct: the R
1 n class contains several R
1 y roots and one R
1 l root; the R
1 y class does not contain all marked R
1 y roots; and there are both R
1 n and R
1 y roots that pattern like regular roots.
There are many other root alternation classes, some more complex than these but, rather than confuse non-Semitist readers even further, I simply ask them to accept on trust my assurance that

34 the analysis that I have given here for coronal sonorant roots extends to the remaining marked alternation classes as well. Every root in every marked alternation class must be flagged or marked individually for membership in its particular class (and no root can belong to more than one class), much in the way that members of verb conjugation classes are flagged on other languages (Aronoff 1994). Most importantly, the fact that these alternation classes in Hebrew operate over individual roots and not lexemes or stems demonstrates the linguistic reality of
Hebrew verb roots, even though these roots have no constant meaning. In the end, Hebrew roots are much like Latin roots. Their patterns are much more complex but they are also much more pervasive: no Hebrew verb form can exist outside the matrix determined by the interplay of root alternation class and binyan. Even the forms of a verb derived from a foreign borrowing must fall into one of the cells determined by the matrix, resulting in such celebrated forms as tilfen ‘he telephoned’ or hizdangef ‘he strolled down Dizengoff Street.’ The form of a Modern Hebrew verb form is completely predictable, though largely unmotivated by either syntax or semantics.
There is one more kind of pattern that distinguishes among the root alternation classes under discussion, their distribution among the binyanim. Although there are seven binyanim, very few roots occur in all seven. Table 13 is based on a tabulation of all the types of verb roots that we have discussed so far, except for the R
1 n roots that pattern like full roots. It includes all roots of these alternation classes that occur in either the hif`il or nif’al binyan and compares their distribution in other binyanim. What is especially interesting is the distribution of hif`il/huf`al or hif`il/nif`al pairs. Both huf`al and nif`al can be passive in meaning and a given hif`il verb, which is always active and usually transitive, is usually paired with one or the other, though huf`al is the more common pairing, since it is the structurally appropriate passive form for hif`il.
It is therefore remarkable that the two largest alternation classes under discussion here, missing

35
R
1 n roots and missing R
1 y roots, differ markedly from one another: the missing R
1 n class favors huf`al overwhelmingly, as expected of all roots, but the missing R
1 y class favors neither huf`al nor nif`al, which is very unusual. Of the remaining two varieties of R
1 y roots, the one that I have grouped together with missing R
1 n roots has the same distribution as that class, while the one that I have classified as full shows no huf`al or nif`al forms at all, again very curious. The overall distribution clearly supports the grouping of the root varieties into traditional alternation classes that I have suggested. Additionally, it suggests another intriguing path to explore, which
I do not follow here, for lack of space: the possibility that root alternation classes differ generally in the way they distribute themselves among the binyanim.
INSERT TABLE 13 ABOUT HERE
There are no other missing R
1
sonorant roots in the language besides the ones I have discussed. I have grouped together in a single alternation class the missing R
1 n roots and the missing R
1 y roots that follow the same pattern as missing R
1 n roots. There is also one missing R
1 l root ( lqx ).
I conclude as follows: i.
There is an alternation class of missing R
1 coronal sonorant roots whose R
1
ranges over all coronal sonorants, [n, l, y], all of whose members must be lexically marked, which includes the traditional missing R
1 n alternation class. ii.
All remaining R
1 n roots, those that do not drop the n , have no lexical marks.
We are left with two varieties of R
1 y roots, those that we may now call true missing R
1 y roots, and those that behave much like full roots. If we agree that no full roots should be lexically marked, then (true) missing R
1 y roots must be marked lexically. We now have two marked alternation classes of missing sonorant-initial roots: missing R
1 y and missing R
1 coronal

36 sonorant , each characterized by a set of forms distributed across the paradigm. Nothing of what
I have shown depends on the roots that belong to these alternation classes having any constant meaning. Indeed, the important generalizations are not about individual roots but rather about root alternation classes. These form a system of conjugation alternation classes and, like all good conjugation systems, have no discernible motivation of any sort but are purely arbitrary grammatical categories. Except that the marked root alternation classes each contain a subset of the roots that begin in coronal sonorants, none of this system can be reduced to phonology either.
Instead, root alternation classes are identified by their paradigms.
xxix
This claim should also be amenable to experimental verification. Modern Hebrew verb roots and their alternation classes, like those of English and Latin, thus furnish yet another example of morphology by itself, leaving lexical meaning to reside where it belongs, not in roots, which are purely grammatical objects, but in lexemes, where language meets the world.
5. Conclusion
My goal in this article has been to show that, though it is difficult to do linguistics while respecting the integrity of words, imposing such a restriction on one’s analyses can lead to interesting results that are otherwise unavailable or unexpected, not just in morphology but also in syntax and lexical semantics. From that general vantage point, we expect that a language newly-arisen with little or no outside influence will have a robust syntax and no sublexical structure at all; we expect that newly-coined words will have idiosyncratic meanings that are understandable only in the extralinguistic context of their creation and whose internal semantic structure is not readily or entirely amenable to syntactically-derived analysis; and we expect to find a language with a very complex matrix of inflected word-forms whose morphological

37 structure is unmotivated in any strict Saussurean sign-based sense. I hope that at least some small fraction of my readers now believe that the kind of work that I have reviewed here is not only possible but also necessary if we are to make progress in understanding language.

38
References
Ackerman, Farrell, and Robert Malouf. 2007. Parts and wholes: Patterns of relatedness in complex morphological systems and why they matter.
Albright, A. 2002. The identification of bases in morphological paradigms. UCLA doctoral dissertation.
Anderson, Stephen R. 1992. A-morphous Morphology . Cambridge: Cambridge University Press.
Aronoff, Mark. 1972. The semantics of growth . Quarterly Progress Report of the Research
Laboratory of Electronics, M.I.T
. 113.159-62.
Aronoff, Mark. 1976. Word Formation in Generative Grammar . Cambridge, MA: MIT Press.
Aronoff, Mark. 1980. Contextuals, Language 56.744-758.
Aronoff, Mark. 1994. Morphology by Itself: Stems and Inflectional Classes . Cambridge, MA:
MIT Press.
Aronoff, Mark, and Kirsten Fudeman. 2005. What is Morphology?
Oxford: Blackwell.
Aronoff, Mark, Irit Meir, and Wendy Sandler. 2005. The Paradox of Sign Language
Morphology. Language 81.301-344.
Aronoff, Mark, Irit Meir, Carol Padden, and Wendy Sandler. 2005. Morphological universals and the sign language type. Yearbook of Morphology 2004 , 19-39.
Blevins, J. 2006. Word-based morphology. Journal of Linguistics 42.531-73
Bloom, Paul. 2000. How children learn the meanings of words.
Cambridge, MA: MIT Press.
Booij, Geert. 2005. An Introduction to Linguistic Morphology . Oxford: Oxford University Press.

39
Carstairs-McCarthy, Andrew. 2002. An Introduction to English Morphology . Edinburgh:
Edinburgh University Press.
Carstairs-McCarthy, Andrew. 2005. Basic terminology. Handbook of Word-Formation , ed. by
Pavol Stekauer and Rochelle Lieber, 5-23. Dordrecht: Springer.
Chomsky, Noam. 1957. Syntactic Structures.
Janua linguarum, no. 4. ’s-Gravenhage : Mouton.
Chomsky, Noam. 1970. Remarks on nominalization. Readings in Transformational Grammar , ed. By A. Jacobs and P Rosenbaum, 184--221. Waltham, MA: Ginn and Co.
Chomsky, Noam. 1995. The Minimalist Program . Cambridge, MA: MIT Press.
Chomsky, Noam. 2000. New Horizons in the Study of Language and Mind . Cambridge:
Cambridge University Press.
Clark, Eve V. and Herbert H. Clark. 1979. When nouns surface as verbs. Language 55.767-811.
Corbin, Danielle. 1984. La formation des mots: structures et interprétations.
Lille: Presses
Universitaires de Lille.
Croft, William. 2003. Typology and Universals . Cambridge; Cambridge University Press.
Downing, Pamela. 1977. On the creation and use of English compound nouns. Language
53.810-42.
Eco, Umberto. 1995. The Search for the Perfect Language . Oxford: Blackwell.
Finkel, Raphael, and Gregory Stump. 2007. A default inheritance hierarchy for computing
Hebrew verb morphology. Literary and Linguistic Computing.
Fraser, Norman, and Greville Corbett.1997. Defaults in Arapesh. Lingua 103: 25-57.

40
Frege, Gottlob. 1914. Letter to Jourdain. Philosophical and Mathematical Correspondence , ed. by G. Gabriel et al. Chicago: Chicago University Press, 1980: 78–80.
Gurevich, Olya. 2006. Construction Morphology: The Georgian Version . Berkeley: University of CaliforniaDissertation.
Hale, Kenneth, and Keyser, S. J. 1993. On argument structure and lexical expression of syntactic relations. The View from Building 20 , ed. By Kenneth Hale and S. J. Keyser, pp.
53-109. Cambridge, MA: MIT Press.12
Harris, Alice. 2004. History in support of synchrony. Berkeley Linguistics Society 13.142-159.
Harris, Zellig. 1951. Methods in Structural Linguistics . Chicago: University of Chicago Press.
Hockett, Charles F. 1960. The origin of speech. Scientific American 203/3.88-96.
Jackendoff, Ray S. 1975. Morphological and semantic regularities in the lexicon. Language
51.639-71.
Jespersen, Otto. 1943. A Modern English Grammar on Historical Principles, Part VI:
Morphology . London: George Allen and Unwin.
Kisch, Shifra. 2004. Negotiating deafness in a Bedouin community. Genetics, disability, and deafness , ed. by John V. van Cleve. 148-73. Washington: Gallaudet University Press.
Lane, Harlan, Richard Piller, and Mary French. 2000. Origins of the American Deaf-world:
Assimilating and differentiating societies and their relation to genetic pattern. The signs of language revisited , ed. by Karen Emmorey and Harlan Lane, 77 - 100. Mahwah, NJ:
LEA.

41
Lees, Robert B. 1960. The grammar of English nominalizations.
Indiana University Research
Center in Anthropology, Folklore, and Linguistics. Publication 12. Bloomington,
Indiana.
Levelt, Willem. 1989. Speaking: From intention to articulation . Cambridge, Massachusetts: MIT
Press.
Malkiel, Yakov. 1967. Each word has a history of its own. Glossa 1.137-149.
Marantz, Alec, and Linnaea Stockall. 2006. A single route, full decomposition model of morphological complexity: MEG evidence. The Mental Lexicon 1.
Meillet, Antoine. 1903. Introduction á l’ étude comparative des langues indo-européennes.
Paris: Hachette.
Nowak, M., & Krakauer, D. (1999). The evolution of language. Proceedings of the National
Academy of Sciences 96.8028-8033.
Sandler, Wendy, Irit Meir, Carol Padden, and Mark Aronoff. 2005. The emergence of grammar: systematic structure in a new sign language. Proceedings of the National Academy of
Sciences 102.2661-5.

42
Saussure, Ferdinand de. 1975.
Cours de linguistique générale . publié par Charles Bally et Albert
Sechehaye, avec la collaboration de Albert Riedlinger ; édition critique préparée par
Tullio de Mauro. Paris : Payot.
Scott, D., R. Karmi, K. Eldebour, G. Duyk, E. Stone, and V. Sheffield. 1995. American Journal of Human Genetics 57 .
965-8.
Spencer, Andrew. 2000. Morphology. The Handbook of Linguistics , ed. by Mark Aronoff and
Janies Rees Miller, 213-237. Oxford: Blackwell.
Stokoe, William. 1960. Sign language structure: an outline of the visual communication systems of the American Deaf.
Buffalo, Dept. of Anthropology and Linguistics, Studies in
Linguistics occasional papers, 8.
Stump, Gregory. 2001. Inflectional Morphology: A Theory of Paradigm Structure . Cambridge:
Cambridge University Press.
Szabo, Zoltan. 2007. Compositionality. Stanford Encyclopedia of Philosophy . http://plato.stanford.edu/entries/compositionality/.
Tarmon, Asher, and Ezri Uval. 1991. Hebrew verb tables.
Jerusalem: Tamir.
Webster’s third new international dictionary of the English language unabridged.
1986. Ed. by
Philip Babcock Gove and the Merriam-Webster editorial staff. Springfield, MA:
Merriam-Webster.
Xu, Zheng, Mark Aronoff, and Frank Anshen. 2007. Deponency in Latin. Proceedings of the
British Academy 145.127–143

43 i
Psycholinguists, e.g. Levelt (1989), use the term lemma for the purely syntactic and semantic aspects of lexemes and the term lexeme for what linguists would call phonological word. I use lexeme to refer to the entire lexical entry, including syntax, semantics, and forms. ii
For a general overview of compositionality (Frege’s principal) see Szabo 2007. I am not suggesting that all linguists agree on what the units are over which compositionality operates, only that compositionality has a role in language. iii
The form pos-it is the allomorph of the root pose that appears before the suffix
–ion
. iv Chomsky provides a sketch of an apparatus that might account for derived nominals nontransformationally, but that apparatus has been very little explored since, except for X-bar theory, which was designed to express the parallelism between the internal structures of phrasal categories like NP and VP without having to derive one from the other transformationally. v
Decompositionalists have their own burdens, most vexing among which is the general problem of recoverability of deletion (Chomsky 1970, fn. 11). vi
For definitions and extended discussion of the terms lexeme and root , see Aronoff 1994, Ch. 1.
I am using morpheme in the classic sense of Harris 1951, extended following the discussion in
Aronoff 1976, Ch.2. vii
The two orders are not symmetrical, but claims about symmetry in this domain have not stood up well over time (Croft 2003). viii
Better candidates for morphological structure are the size and shape classifiers that are found in some signs in ABSL. These are limited in complexity and productivity, but may be precursors of the rich and productive classifier constructions found in more established sign languages
(Emmorey 2003).

44 ix
Syntactic Structures itself has a prescient short section (7.3) showing that the adjective interesting is not derived from a sentence containing the verb interest , for precisely the sorts of reasons that I outline below. x
One indicator of its popularity at the time is the fact that it was reprinted four times by 1966, the date of issue of my own copy, no mean feat when one takes into account the size of the community of linguists in the early 1960’s. xi
"There's glory for you!"
"I don't know what you mean by 'glory,' " Alice said.
Humpty Dumpty smiled contemptuously. "Of course you don't—till I tell you. I meant
'there's a nice knock-down argument for you!' "
"But 'glory' doesn't mean 'a nice knock-down argument,' " Alice objected.
"When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean—neither more nor less."
"The question is, " said Alice, "whether you can make words mean so many different things."
"The question is," said Humpty Dumpty. "which is to be master—that's all."
Lewis Carroll, Through the Looking Glass xii
Incidentally, an example like skin demonstrates on its face the wrongheadedness of approaches to conversion pairs based on roots or underspecification, in which the meanings of both the verb and the noun are derived from some third category that is neither a verb nor a noun: the direction of the semantic relation is clearly one-way, from noun to verb. The different senses of the verb are all readily traceable to the sense of the noun, but there is no inverse relation for any sense of the verb, so that the verb must be derived from the noun. If we start from some third point, we can’t account for the directionality. The opposite derivation, from verb to noun, holds for instance or result nouns like hit . xiii
The verb friend never appears in official Facebook text, only among users. xiv
These examples were supplied by my daughter Katy.

45 xv
I am not denying that non-causative individual verbs can have multi-predicate meaning. The verb propose ‘to make an offer of marriage’ (WIII) is similar to friend in involving an invitation.
The question is whether there is a systematic multi-predicate syntactic relation between nouns and their derivative zero-verbs involving conative or similar aspect. xvi
I say M
ODERN
H
EBREW
because the transcription in this article is of Modern Hebrew, but in fact the patterns I discuss are largely identical in the earliest attested forms of the language. I use
B
IBLICAL
H
EBREW to refer to the language of the unpointed text of the Bible, dating between about 800 and 100 BCE. I use M
ASORETIC
H
EBREW
to refer to the pointed text of the Bible, dating from about 600 CE. xvii
Earlier stages of Hebrew or related Semitic languages do not show much more regularity in this regard. xviii
Even these examples are a little forced, since some of the lexical items are very infrequent, obsolete, or even artificial. In the early days of Modern Hebrew, linguistic zealots coined numerous words that never caught on; many of these survive in dictionaries and only a sensitive native speaker can distinguish them from words in everyday use. xix
I confine my discussion to verbs, both because there are fewer verb patterns than noun patterns and because it is easier to believe that verb roots are related semantically, so if we can talk about verb roots without recourse to meaning, then all the more so for nouns and adjectives, whose meanings show much greater variety. xx There are roots with more than three consonants, but these are not germane to our conversation.

46 xxi
Not even the most abstract of phonological representations could save the Hebrew system from being morphologized. In the case of n , for instance, one would have to posit n
1 and n
2
, without any phonetic way to distinguish them beyond the fact that they pattern differently. xxii
Finkel and Stump (2007) adopt a similar approach in their computational approach to Hebrew verb forms. xxiii
There are no longer any geminate consonants in the language, but former geminates are still exempt from spirantization. xxiv In Modern Hebrew, the corresponding form is nipol . xxv
The root ngd is ancient, but its use in verbs is not, which shows that what exempts a verb root from assimilation of n is not the age of the root, but rather the point at which it was first used as a verb. xxvi
The failure of b to spirantize to v in the first example shows that there is a missing n before the b in the root of hibit . xxvii
All three R
1 y roots in which the hif`il forms follow the missing R
1 n pattern are also R
2 c .
Two other R
1 y R
2 c roots do not occur in hif`il, but their pa`al forms follow the missing R
1 n pattern ( ycq and ycr ), supporting the observation that R
1 y R
2 c roots follow the missing R
1 n pattern generally. However, the one other root of this form, yc’
, falls into the local default R
1 y class in hif`il ( hoci’
, yoci’
). This sort of distribution, in which a circumscribed exception to an exception follows a local default rather than a general default, is not uncommon (Fraser and
Corbett 1997), and supports the assignment of these five R
1 y R
2 c roots to the missing R
1 n class through a very local phonological generalization. xxviii
These roots also share the less common root vocalism a in the future and imperative stem.

47 xxix
There has been a good deal of recent work in in which morphology is viewed primarily as a complex system of word-forms (e.g. Albright 2002, Harris 2004, Blevins 2006, Gurevich 2006,
Ackerman and Malouf 2007, Stump and Finkel in press). This other work does not deal specifically with the morphology of roots and the theoretical perspectives are by no means uniform, but none of them is morpheme-based in the traditional sense.

Table 1: ABSL is predicate-final

Table 2: Major constituent orders in ABSL data
49

Table 3: relative orders of a modifier and its head in ABSL
50

51
Locatum verbs
Locatum verbs
A.
On / Not on
B.
In / Not in
C.
At , to
D.
Around
E.
Along
F.
Over
G.
Through
H.
With
LOCATION AND DURATION
VERBS
LOCATION VERBS
A.
On / not on
B.
In / not in
C.
At , to
DURATION VERBS
AGENT AND EXPERIENCER
VERBS grease spice poison frame hedge bridge tunnel trustee
Goal and source verbs
Goal verbs
A.
Human roles
B.
Groups
C.
Masses
D.
Shapes
E.
Pieces
F.
Products
G.
Miscellaneous
Source verbs
INSTRUMENT VERBS
A.
Go
Ground
Lodge
B.
Fasten
C.
Clean
Dock D.
Hit
Summer E.
Cut , stab
F.
Destroy
G.
Catch
H.
Block
I.
Follow
J.
Musical instruments
K.
Kitchen utensils
L.
Places
M.
Body parts
N.
Simple tools
O.
Complex tools
P.
Miscellaneous
MISCELLANEOUS
VERBS widow gang up heap loop quarter nest cream word
Bicycle
Nail
Mop
Hammer
Harpoon
Grenade
Trap
Dam
Track
Fiddle
Ladle
Farm
Eyeball
Wedge
Mill
Ransom

52
AGENT VERBS
A.
Occupations
B.
Special roles
C.
Animals
EXPERIENCER VERBS
Butcher
Referee
Parrot
Witness
A.
Meals
B.
Crops
C.
Parts
D.
Elements
E.
Other
TABLE 4: CATEGORIES OF ZERO-VERBS FROM CLARK & CLARK 1979
Lunch
Hay
Wing
Snow
House/ s/

Word-Final Root Sample verb sume resume mit pel permit repel ceive duce scribe pete/peat cur receive deduce prescribe compete recur
Root +ion resumption permission repulsion reception deduction prescription competition recursion
Root +ive resumptive permissive repulsive receptive deductive prescriptive competitive recursive
TABLE 5: Some Latinate Roots in English with their alternations
53

ori tue f(or) hor(t) luct meti fate prec quer apisc fru fung medit spic ut
Latin deponent root gradi la:b sequ min loqu nasc moli mori
Number of verbs with such a root
22
16
15
11
10
10
9
9
7
7
5
5
8
8
7
7
4
4
4
4
5
4
4
Example verb luctor metior fateor precor queror apiscor fruor fungor meditor conspicor utor gradior la:bor sequor comminor loquor nascor molior morior orior tueor affor hortor
Gloss wrestle measure concede ask for regret grasp enjoy perform contemplate see make use of proceed glide follow threaten talk be born build up die rise look at address encourage
TABLE 6: Latin deponent verb roots
54

Root = zlp
Nouns zelef zalfan mazlef hazlafa ziluf zlifa
Verbs zalaf zilef hizlif zulaf huzlaf
Adjective mezulaf
‘sprinkle, spray, drip’
‘sprinkling fluid (perfume)’
‘sprinkler’
‘watering can, sprayer’
‘sprinkling’
‘sprinkling’
‘sprinkling’
‘to pour, spray, sprinkle’
‘to drip’
‘to sprinkle’
‘to be sprinkled’
‘to be sprinkled’
‘sprinkled’
Table 7: lexemes formed on the Modern Hebrew root
ZLP
55

Root = skr
Nouns saxar soxer saxir sxirut sxira haskara histakrut maskir maskoret
Verbs saxar niskar histaker hisker huskar
Adjectives saxur musxar niskar
‘pay, rent, lease’
‘hire, wages, profit’
‘tenant’
‘hired laborer, employee’
‘rent’
‘hiring, renting’
‘lease’
‘wages, profit, renting’
‘landlord’
‘salary’
‘to hire, rent’
‘to be hired, rewarded, paid’
‘to earn wages, make profit’
‘to lease, let’
‘to be leased’
‘rented’
‘let, leased’
‘hired, benefited’
Table 8: lexemes formed on the Modern Hebrew root SKR

56

/s/ is the Hebrew letter sin .

Root = kb∫ keve
∫ kvi
∫ kvi
∫ a kiv
∫ on maxbe∫ mixba∫a
Verbs kava∫ kibe∫ hixbi∫
Adjectives kavu∫ kvu∫im mexuba∫
‘press’
‘gangway, step, degree, pickled fruit’
‘paved road, highway’
‘compression’
‘furnace, kiln’
‘press, road roller’ pickling shop
‘to conquer, subdue, press, pave, pickle, preserve, store, hide’
‘to conquer, subdue, press, pave, pickle, preserve’
‘subdue, subjugate’
‘subdued, conquered, preserved, pressed, paved’
‘conserves, preserves’
‘pressed, full’
Table 9: lexemes formed on the Modern Hebrew root
KB
∫
57

Root ntp ncl npl npl npl nbt ntn ntn lkx npc npk ngd
Binyan ntp pa`al hif`il nif`al pa`al hif`il huf`al hif`il pa`al nif`al pa`al pa`al pa`al hif`il
Past nataf hi-tif ni-cal *
(cf. nixtav) nafal hi-pil
Future yi-tof
Infinitive li-tof
Gloss drip
(intr.)
Remarks n drops before
R
2
( t ) ya-tif le-ha-tif drip (tr.) yi-nacel le-hi-nacel arrive *would be nincal if n were not assimilated yi-pol ya-pil li-pol fall le-ha-pil cause to fall hu-pal ju-pal n.a. be thrown down hi-bit* ya-bit* le-ha-bit* look natan nitan*
(cf. nixtav) yi-ten la-tet* give
*b would be v if n were not assimilated
*infinitive is irregular, should be liten yi-naten le-hi-naten be given *would be nintan if n were not assimilated lakax yi-kax la-kaxat take nafac nafak hi-gid yi-poc yi-pok ya-gid linpoc linpok disperse, smash go/come out, result le-ha-gid tell only example of initial l that behaves like missing R
1 n n does not assimilate in infinitive n does not assimilate in infinitive
Table 10: missing R
1
n roots
58

npk ngd nbt ngb ngb ngb ncx hif`il pa`al hif`il pa`al pi`el hif`il hif`il hi-npik ya-npik le-hanpik nagad hi-nbit nagav yi-ngod ya-nbit le-ha-nbit sprout yi-ngov yigov
, li-ngod li-ngov issue oppose become dry n drops variably in the future nigev ye-nagev le-nagev dry hi-ngiv ya-ngiv le-ha-ngiv go south hi-ntsíax ya-ntsíax le-hantsíax memorialize, eternalize
Table 11: full R
1
n roots
59

Sample root nmx nzq yld ycg (only 3 roots) ynq (only 4 roots)
Root Variety full R
1 n
Past 3m.sg. hi-nmix missing R
1 n missing R
1 y default, prefix vowel = o missing R
1 y , like missing
R
1 n full R
1 y , prefix vowel = e hi-ziq ho-lid hi-cig he-yniq
Future 3m.sg. Gloss ya-nmix lower, humiliate ya-ziq yo-lid ya-cig ye-yniq
Table 12: R
1
n roots and R
1
y roots in hif`il damage beget show, exhibit, present suckle, water, saturate
60

occurring binyan combinations missing R roots hif’il/huf’al/nif’al 4
1 n hif’il/huf’al
19
1 hif’il/nif’al huf’al/nif’al hif’il only nif’al only
TOTAL
1
5
9
39
R
1 y default roots
2
7
8
0
7
5
29
R missing R
0
3
0
0
0
0
3
1 y like
1 n
R
0
0
0
0
4
0
4
1
Table 13: Root Type and binyan distribution (R
1
n/y) y full roots
61

Figure 1a KETTLE (CUP + POUR)
Figure 1b: KETTLE (CUP + POUR)
62

63
FIGURE 2a and 2b: BANANA

FIGURES 2c, d, and e: BANANA
64