You are on page 1of 34

The Linguistic Review l :81-114

Argument Structure and Morphology


EDWIN WILLIAMS

That there is some regulr relation between the argument structures of morphologically related words is clear and well-known, and there are several proposals about how such relations can be characterized. This paper presents another such proposal, one that differs from previous proposals mainly in that it represents an attempt to be highly restrictive, perhaps erring in that direction. By argument structure of a lexical item I mean simply a labelled listing of the arguments that a lexical item can have. I will adopt the labelling proposed by Gruber (1976) (Actor, Theme, Goal, Source), though the actual labels themselves are not important. These are called thematic relations. Now suppose that a lexical item has an argument structure, and we apply some morphological rule to that lexical item to derive a new lexical item, with a new argument structure. What we want to characterize in this paper is the function that will relate the old argument structure to the new argument structure. One way to do this would be to dive right in and write a notation for writing argument structure transformations in. This is essentially what past proposals have done. What I intend to do instead is to specify two functions ("internalize X" and "externalize X"), each parameterized by the thematic relations. Since there are finitely many thematic relations, there are finitely many functions, and the claim is that these exhaust the possibilities. Before these functions can be specified, there are a number of topics that must be discussed which are independent of these functions. For example, we must have a theory of exactly how many argument types (thematic relations) there are. Furthermore, we must determine whether there is any further structure to the argument structure of a verb than
* This paper is a revision of a paper with the same title written in February, 1980. The author wishes to thank the Max Planck Institute for Psycholinguistics for the opportunity to complete the work on this paper s a member of its research staff in the summer of 1980.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

82
that embodied in the notion "unordered list of arguments". In Section l, I will propose that there is one minimal additional element of structure to the <4unordered list of arguments": the distinction between "internal" and "external" arguments. Such things would need to be studied even if there were no morphology or no need for any function from one argument structure to another. We also would need some specification, independent of morphological functions, indicating how the various arguments, or thematic relations, are "realized" in syntax; that is, for a given verb in a given sentence, how do you know which phrases in that sentence correspond to which arguments of that verb. There are innumerable ways to specify the realization of the arguments, and there are several important issues. One issue concerns the need for grammatical, in addition to thematic, relations. Without grammatical relations, it is necessary to specify directly the realization of each thematic argument; for example, that the theme is realized s the NP directly dominated by the VP that dominates the verb, etc. But if there are grammatical relations, this specification can be indirect; for example, one might specify that the theme is the direct object and that the direct object is the NP immediately dominated by the VP that dominates the verb. I will explore the possibility in this paper that there are no grammatical relations; that is, that the realization of the thematic relations is directly specified. Some of the issues involved in this decision are the familir ones: if there are no grammatical relations, then of course there can be no rules, either syntactic or lexical, that refer to them, etc. Another issue that arises in the "realization" problem is the level of generality that holds for the rules governing the syntactic realization of arguments, and similar problems arise in theories with and without grammatical relations. In the worst case, every verb of every language would need to specify the realization of every one of its arguments - and in the best case, realization would be completely fixed, admitting of no Variation. Now reality clearly lies between these: given that some languages realize arguments positionally and others according to a case System, and others with some mixture of these, then the best case is not possible. On the other hand, it seems possible to avoid specification of realizations for particular lexical items in particular languages, so the worst case does not hold either. Now it appears that some parameters of realization can vary from language to language, and others from category to category within a language, though hopefully the latter can be kept to a minimum. In Section 2 we will specify a solution to the realization problem for English and make some general remarks about the universal characterization of realization rules. Now, the material in the first two sections, s I mentioned earlier, is independently required; even in a language with no morphology, we will

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

83
want to know what the argument structure of lexical items is, and how arguments are syntactically realized, and we will want universal theories of these things. Clearly, how these theories are formulated will determine to some extent what kind of relation can hold between the argument structures of different morphologically related words. It is to this that Section 3 is devoted.
1. ARGUMENT STRUCTURE

As mentioned above, we will take the argument structure of a lexical item to be an unordered list of its arguments, with one further complication. We will use the labels proposed by Gruber (1976), whichare Actor, Theme, Goal, Source, etc. Thus we might represent the argument structure of hit in the following way: (1) hit: (Actor, Theme)

Now, in the theory we will develop here, we will want to distinguish one of the arguments in an argument list, which we will call the "external" argument, and we will adopt the notation of underlining to mark this argument; thus the argument structure of hit might be: (2) hit: (Actor. Theme)

The notion external argument corresponds in some respects to subject, but we avoid that term because of its ambiguity. The external argument is simply a distinguished argument, not a syntactic position, a case, or something eise. Another reason we avoid the notion subject is that this term implies other terms s well, such s direct object, indirect object, etc., which will play no role in our svstem. The obvious question at this point is: why is there one and only one external argument in an argument structure? We can answer this question with reference to the theory outlined in Williams (1980), where the term external argument was introduced. In that paper, it was proposed that the subject predicate relation was to be indicated by coindexing subject and predicate; one reason for this was that a c-command restriction on coindexing then would automatically apply to the subject predicate relation, requiring that a subject c-command its predicate, and this seems correct. Thus for example, the subject of a sentence and its verb phrase will be coindexed by a rule of predication:
(3) [

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

84
The use of indices to mark the subject predicate relation has another important consequence. If we take the X-bar theory seriously, then every phrase, and in particular, every predicative phrase, apart from S, is the maximal projection of some lexical item. The use of indices to indicate the subject predicate relation entails that there can be only one subject for any maximal projection used predicatively, since any maximal projection can bear only one index. In essence then, all predicative phrases that are maximal X-bar projections are necessarily monadic. Given that lexical items are not monadic in general, but that maximal projections of them are, we may ask what the relation is between a head and its maximal projection. Clearly, all of the arguments of that head must be specified internal to the maximal projection, except for the one external argument that is, the item that is located external to the maximal projection but with which the maximal projection is coindexed. Thus, for example, the verb A/Y, s we mentioned, has two arguments: an Actor, which is external, and a Theme, which is internal. This means that in a sentence with hit, the theme must be specified within the verb phrase of which hit is the head, and the Actor of hit must be specified in a position external to the verb phrase of which hit is the head, but with which that verb phrase is coindexed. Thus, the external argument of a lexical item is located outside of the maximal projection of that item, and since it must be coindexed with that maximal projection, and only one index is allowed, there can only be one external argument. There is certainly no necessity forthings tohave worked out this way, and in fact, if one adopts an unconstrained use of the lambda calculus, s was done in Williams (1977), then things will not work out this way. If a maximal syntactic projection receives a logical translation which lookslike: (4) [XxXy....]^

then the thesis is denied: this phrase has two external arguments, one corresponding to each lambda. In essence then, the use of coindexing to indicate the subject predicate relation amounts to a limited use of the lambda calculus: you only get one lambda per X-bar maximal projection, and the index on the predicative phrase serves the role of this single abstraction operator. So, the external argument of a lexical item is the one that corresponds to the NP in a sentence of which a phrase with that item s its head is predicated. The use of indices to indicate the relation between a predicative phrase and its subject sets an upper bound (of one) on the number of external arguments that that phrase can have, but no lower bound. The lowest conceivable bound is, of course, zero, and this is the analysis given

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

85

in Williams (1979) to the VP with head seems. Bresnan has shown that seems cannot have a "subject" (Bresnan 1972): (5) *That John is here seems *John's presence seems

In the terms outlined here, this amounts to the claim that there can be no external argument. We can indicate this by giving seems an argument structure in which nothing is underlined: (6) seems: (Theme, Goal)

This means that when seems heads a VP, all of its arguments must be located internal to that VP; hence, seems has no external argument. In order to maintain this, we must distinguish between occwrence ofan argument and binding of an occurrence of an argument. An item might occur internal to a phrase but be bound external to the phrase; it is nevertheless an internal argument of the head of that phrase. This is exactly what happens with raising seems: (7) Johnj [seems [tj to be happy] to me.Jyp.

Here, both of the arguments of seems (the Theme, t to be happy, and the Goal, to me) occur within the VP. However, part of one of the arguments of seems is bound from without: John binds the trace. In this case,/0A/z is the subject of the predicative phrase "seems to be happy to me" but not by virtue of corresponding to the external argument of the head of that phrase, which does not have one, but rather by virtue of binding a trace in that VP. Of course, phrases other than VP are used predicatively. APs for example, s in John is proud of Mary. Here, proud has two arguments, one internal, Mary, and one external, John. There are also adjectives like certain, which have no external argument; hence we have raising for this kindof adjective: (8) Joluij is [certain [t. to win] ]yp i (Theme)

The case of Noun Phrase is a bit more complicated, but the result will be the same: that nouns have one and only one external argument. In the case of verb, when there is an Actor and there is an external argument, then the Actor is the external argument - this is a fairly clear universal. Now, nouns have Actors s well; witness the famous NP: thearmy's destruction

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

86
of the city, or its passive: the city 's destruction by the army. But note that in both of these, the Actor is internal, that is, it is located inside the maximal projection of N (NP, or N"). So, if Ns have external arguments, they are not Actors. But do Ns have external arguments? I believe that we want to say that they do, at least when they occur s the head of a NP that is used predicatively, s in "John is a fool". We might say that fool had an external Theme (that is, it has the argument structure (Theme). Then we could say that when fool appeared s the head of a NP used predicatively, the subject of that predication corresponded to the theme of fool. But we can show that this will not do generally, because we can show that this would require some nouns to have both internal and external Themes. Consider the following, forexample: (9) I consider that [destruction of a city by evil forces]

Here, "destruction of a city by evil forces" is predicated of "that". But the predicative NP has internal Theme and Actor - thus to what argument of destruction does that correspond; i.e., what is the external argument of destruction, if not Theme or Actor? Clearly the external argument of such a noun has no counterpart in the verbal System; suppose we invent the label R to name that argument of the noun which is external. Then we would assign destruction the argument structure (R, Actor, Theme). The label R is meant to suggest "referential", since it is this argument position R that is involved in referential uses of NPs s well. Thus, for example, if fool has the argument structure (E), we might assign the following logical representations for the predicative and referential uses respectively: John and the variable is in the position of R: (10) John is a fool The fool left fool (John) E!x(fool (x) & left(x) )

To sum up this section, we have assigned a minimum amount of structure to the noon argument structure. The argument structure of a lexical item is simply a list of its arguments. In addition, one of those arguments is distinguished from the rest s being the external (underlined) argument. The external argument of a lexical item corresponds to the NP of which the major projection ofthat lexical item is predicated. It is worth emphasizing, that for our purposes the argument labels (Actor, etc) have no intrinsic content or structure; the only property that each one has is that it is different from the others. No doubt "Actor" has some cognitive content, but we claim that this is linguistically irrelevant; i.e., that the rules determining the argument structure of derived words

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

87

which we will discuss in Section 3 are not concerned with whatever that content is.
2. THE SYNTACTIC DEPLOYMENT OF ARGUMENTS

We now tum to the characterization of the rules of "realization". The basic question, for a lexical item in a sentence, is how is it determined which NPs in that sentence function s realizations of which arguments of that lexical item? A partial answer has already been given: the internal arguments must be realized internal to the maximal projection of the lexical item, and the external argument (if there is one) must be coindexed with the maximal projection of the lexical item. The coindexation of the external argument is governed by the rules of predication, s outlined in Williams (1980), and will not be the subject of further discussion here. In the previous section t wo universals concenng the external argument were given: (11) l. If there is an Actor, it must be external for V 2. R must be external for N

The first of these will have important consequences for the analysis of the passive constructioii, which we wfll consider in Section 3. We now turn t o the deployment of the internal arguments. To describe the deployment of the Theme of the verb Hit, for example, we might say, "the Theme of h is the NP that is immediately dominated by the VP of which hit is the head" - or, to use a familir notation: (12) Theme: (NP,VP hit )

Similarly, to describe the location of the Goal of a verb like give we might write: (13) Goal:(NP,<PP to ,VP,*e))

In fact, though, the reference to VPhit and VP^ is entirely superfluous, since by definition these arguments must be internal to the maximal projection of the lexical item (in this case, V) of which they are arguments, so we may drop the reference to VP. We might suppose s well that these particular realizations are not particular to the verbs in question, but in fact are valid for verbs that have the argument types Theme and Goal, and thus we may drop the reference to particular verbs. We are left with general realizations for Theme and Goal:

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

88
(14) Theme: NP Goal: (NP,PPto)

The subscripting notation X will be useful in case languages, where, for example, we will refer to realizations like the following:

(15)
which is to be read: the Goal is the NP headed by the feature "dative" (or, whose head bears the feature "dative"). Notice that the realization for Goal given above is not category specific, since we have dropped the reference to VP. This is a welcome result, since Goal is realized s PPto in all English categories: (16) the trip to Houston apparent to Mary

In general, a realization rule wl have the form: (17) A:(X y ,Z w )

where X and Z are major categories, and y and w are either prepositions or case features, and A is one of the argument types. We will assume that the realization rules are verb-independent; that is, if a verb has an argument of type X, then any realization rule for type X in the language will be applicable. Below, we list some of the realization rules for English: (18) Actor:(NP,PP by ) Actor:(NP,PP of ) (limited to NP) Goal:(NP,PPto) Goal: (NP2) Theme: (NP) Theme :(NP,PPof) John was seen by Bill The destruction by the army The shooting of the hunters to give to John the gift to John give Bill the book hitBill the destruction of the city deprive Bill of money aware of the accident speak of something arrive from Houston the arrival from Houston

Source:(NP,PPfrom) X: (NPposs)

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

89

Most of these are self-explanatory. The last one says simply that a possessive NP may bear any relation whatever to the head noun; this is a great exaggeration, but it is a first approximation that is difficult to improve upon. The subscript "2" on NP in the realization rule for Goal is simply a shorthand for whatever the appropriate distinction is between the first and second NP of the double object construction in English; perhaps they differ in terms of a case distinction that is not visible in English. It is important to emphasize that these rules are specified once and for all for a language (English, in this case) and that no particular lexical item can set up its own realization rules. Thus this Information is not listed with particular lexical items. It might be objected here that our claim to have avoided the use of grammatical relations is radier empty now, since the realization rules we have offered are written in the notation that was originally invented (in Chomsky (1965) ) for grammatical relations. But this objection misses the point. In a theory with grammatical relations, there will be two sets of entities, the thematic relations and the grammatical relations, and there will be two sets of realization rules - one set which says which thematic relations are realized s which grammatical relations, and another set which says which grammatical relations are realized in which syntactic configurations. In the theory here, there is one set of realization rules and one set of entities. There is more at stake than these numbers, however. Given that the set of thematic relations and the set of grammatical relations is not isomorphic (under the realization rules), then in a theory with grammatical relations, one might imagine there being rules, say morphological or lexical rules, which operate on representations, say of words, purely in terms of grammatical relations, s well s rules that operate on thematic relations. Of course this will be impossible if there are no grammatical relations. In the system outlined in this paper, the closest thing to a grammatical relation is the distinction between internal argument and external argument, where external argument corresponds to the notion subject. Even here, though, there is a great difference. In the Systems with grammatical relations, the subject argument has the same Status s the others; its realization, for example, is written in the same notation (e.g., in the System of Chomsky (1965), the subject is (NP, S) whereas the object is (NP, VP)). In the System of this paper, however, the two receive entirely different treatments. The internal arguments are subject to the realization rules; the subject, or external argument, is handled by an entirely different mechanism, the predication coindexing mechanism outlined in Williams (1980). Thus, although some notation has been borrowed from the definition of grammatical relations in Chomsky (1965), the system in this paper differs in a number of important respects from Systems based on Brought to you by | Chinese University of Hong Kong (Chinese Univers grammatical relations.
Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

90

There are some problems to which we will return for detailed discussion in Section 6 with the notion that the realization schemes are not verb-particular. This predicts, for example, that if there are two ways to realize a certain argument type, then a verb that has an argument of that type will be able to use both realizations. And there are many examples in which this is true; for example, many verbs with Goal can use the double object construction, s well s the PPto construction. But what about the verbs for which only one of these is possible (donate, for example)? For these we can use the mechanism of subcategorization: donate is NP PP, whfle give is NP NP and _NP PP.
3. HOW MORPHOLOGY CAN AFFECT ARGUMENT STRUCTURE

It wfll be our contention that a morphological rule can affect only the external argument of its input, and that it can affect this external argument in only one of two ways: it can make one of the internal arguments into the external argument, or it can add a new external argument. Obviously, morphologically related words differ in more ways than this will be able to account for. For example, the subcategorizations of morphologically related words differ, and when the words are in different categories, the realizations might be different. We will first review some of these differences, suggesting that they in fact need not be specified in any particular morphological rule, but will follow from general considerations. Afterwards, in Section 3.2, we will examine the actual morphological rules themselves. 3.L Predictable Side-Effects of Morphological Rules Morphological rules do not specify a number of observable differences between their inputs and their Outputs. Consider, for example, the rule of nominalization which derives hatred from hate. Both of these words have internal themes: the hatred ofBl, I hate Bill. In the case of the verb, the theme is realized s a direct object, but in the case of the noun, in a PPof. Why this difference? Now, a verb can have its theme in the direct object, or in a PPof (s in to speak of x), but a noun can have only the latter, never *hatred Bill. But surely this is because the base rules of English do not generate NPs in the complement structure of NP (or perhaps that the case rules do not assign case there); and since there is only one other realization of Theme, PPof, NP must use that one. Adjectives also used PPof for Theme, and presumably for the same reason s in the case of NP. The point is that none of this need be specified by the rule which takes hate into hatred - the difference wfll follow automatically from other differences between NP and VP.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

91

In the above analysis of the nominal of, we make no recourse to an of insertion transformation, so popul r in studies of nominalizations. We have several reasons for not doing so. First, in nonthematic objects of V, we do not get of in the corresponding nominal: (19) John's promise r of Pete to leave. Uo J

We know that the direct object of promise is not a theme because it does not control the complement clause (see Williams (1980) ) and because it does not Heavy NP Shift: (20) *John promised to leave all of the people that were there. (cf. * John promised the books all of the people who were there.)

A theory with of insertion does not predict this; the scheme we have outlined earlier does thoughonly Themes can be realized in PPof. In sum then, the fact that nominalizations of verbs with direct object themes always have PPof themes will follow from features of the base (or of case theory, which might account for this difference in the base) and need not be specified in morphological rule. Subcategorizations of mohologcally related words differ s well. For example, readable differs from read, from which it is derived: (21) read: _NP readable: .

Again we can ask whether this difference is specified in the morphological rule which relates these two. Hopefully, it will turn out that it is not necessary to specify it for each rule; hopefully, it will follow from an examination of those differences which are assigned to related lexical items by mohological rule. We will return to this question in Section 6. 3.2. TheRules Postponing until later further questions about what morphological rules need not specify, we turn now to a delineation of that part of the rules of morphology which is specified in each rule and which concerns the mapping of the input argument structure into the output argument structure. Our claim is that morphological rules can alter argument structure in only two ways. To preview, a rule can efther (a) externalize an internal argument, or (b) internalize the external argument. We will symbolize these s:

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

92

(22)

E(X),I(X)

where X is the name of one of the thematic relations. We wfll define these operations in 3.2.1 and 3.2.2 respectively. These two ftmctions specify only the change of argument structure between the input and output of a morphological rule. Obvioudy, morphological rules specify other changes; for example, the phonological form of the output (in terms of the input), the category of the output, etc. These other aspects of morphology will not concern us here. We may assume any of a number of different views on these matters (see Aronoff (1976); WiUiams (1981); or Allen (1978) ); the choice will not matter for our present purposes. Whether a morphological rule is a rule of affixation or whatever, it has an input stem or word and it defines an output word. The input word or stem has an argument structure, and the morphological rule specifies the Operation on that argument structure that will derive the argument structure of the output word. We are claiming that the Operation must be one of the two mentioned above. Clearly, the argument structure of the output of a morphological rule must conform to whatever well-formedness conditions apply to the argument structures of words in general, or of words in that language, or of words in the category of the output word, etc. For example, if the output of the rule is Vs, and if there is an Actor in the output, then the Actor must be the external argument if there is one; this, because there is a universal law to this effect, and this law, like all laws, holds for both mo hologically complex and morphologically simple words. Similarly, if the output is a noun, then the external argument will be R, unless the rule specifies otherwise. All this is s expected, if these conditions apply to words. 3.2.1. E(X) - "Externalize an Argument": E(X) is an Operation (or rather a set of operations parameterized by the set of thematic relations) on argument structures, which yields new argument structures. The Operation is defined s follws: (23) E(X): erase the underline on the external argument, if there is one, and underline X. If X=0, then underline nothing.

This Operation takes an internal argument and makes it external, where X is the intemal argument that is to be made external. If X is not specified in some particular instance of this rule, then nothing is made external, and so we derive words with zero extemal arguments. If X is specified in some instance of this rule, and if some word does not have an X, then the rule is not defined for that word.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

93

We will now look at several examples of E(X) rules: One instantiation of E(X) is E(theme), or "Extemalize the theme". Two examples of this are the rule which adds -able to form adjectives, and the rule which creates the adjectival past participles. Given a form like read with argument structure (A, Th), and a rule of morphology which (a) adds -able to the stem, and (b) applies E(Th) to the argument structure, we will derive readable, which has argument structure (A, Th). This can be symbolized s: (24) E(Th): read (A, Th) >readable (A, Th)

This correctly represents the meaning ofreadable, and its relation to read: the subject of readable corresponds to the direct object of read. Furthermore, s Wasow (1977) has pointed out, -able can be added only when the Theme is externalized (to use my terminology, not his): (25) Those things are | promisable l (theme externalized). jperishable J *Those people are runnable (Actor externalized). Those people are promisable (Goal externalized).

In Williams (1979), Aronoff (1972) and Wasow (1977), it is shown that there are two passives in English - roughly, an adjectival passive and a verbal passive. These two are most clearly distinguished in Wasow (1978). The adjectival passive permits the attachment of -un (which attaches only to adjectives), it can be fronted by WH-movement (s all adjectives, and no verbs, can) and, most importantly for our purposes, the adjectival passive always has a Theme for its subject, whereas the verbal passive can have Theme, Goal, or Source for its subject. The two passives are illustrated below: (26) Adjectival: The rules are ungiven (Theme of give is subject of given). *We are ungiven (the rules) (Goal ofgive is subject of given). How firmly promised are these things (Theme ofpromise is subject of promised). *How firmly promised were those people (Goal ofpromise is subject of promised). Verbal: Those people were promised these books. These books were promised to those people.

/e will give an analysis of these two passives in terms of the theory of

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

94

morphology developed here. We will assume what is argued for in Williams (1979); namely, that the verbal passive involves NP movement, the classical, or EST assumption concerning the verbal passive, and we will assume that the adjectival passive does not involve trace hinding. The goal, of course, will be to try to explain how this is so, and why it must be the way it is. This goal cannot be achieved solely in terms of the theory of this paper, but in conjunction with the theory of case, we can go fairly far towards this goal. To begin with the adjectival passive, note its similarity to -able affixation. In both cases we derive an adjective, and in both cases we have thematic constancy: the Theme must be the subject of the output word. From this we conclude that the adjectival past participle is derived by E(Th), just s adjectives in -able are. The adjectival passive, then, has no interesting syntactic component to its description - no NP movement, for example beyond the fact that the adjectival past participle is an adjective; to syntax, it is just another adjective. The verbal passives, on the other hand, cannot be derived by any morphological rule (alone), however, since the verbal passives do not exhibit thematic constancy; that is, the subjects of verbal passives are not uniformly of one thematic type (we have seen that the verbal passive can have at least Theme and Goal subjects). Thus the verbal passive cannot be analyzed s E(X), for any particular X. On the other hand, though, the argument structure of the verbal passive cannot be simply identical to the active either, for we would then have no account of the difference between the active and the passive. In the following, we will entertain the idea that the verbal passive participle is formed by the Operation of E(0) (or just E( ) ), which we may read s: (27) E(0): erase the underlining in the argument structure giving an argument structure with no external argument.

In other words, the verbal passive has no external argument. There is already a class of underived words with this property, s we noted earlier: the verbs seem, appear, etc. So E(0) simply maps verbs into this class. (See (5) ). Unfortunately, this proposal makes quite peculiar predictions for the verbal passive; it predicts, for example, the grammaticality of the following: (28) *It was seen Bill by Fred. E(0): see (A,Th) > seen (A, 1h)

Here, all of the arguments of the verbal past participle are internal to the

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

95

VP, and the dummy fills the subject, with which no argument is associated. Now, in syntax, there is a rule of NP movement, and if we apply it to this example, we can derive the grammatical: (29) Billj was seen t{ by Fred. (A,Th)

Here, all of the arguments ofseen (t{ and Fred) are again internal; s in the case of seem raising, we must distinguish between external occurrence of an argument and the binding of an internal argument by an external NP.1 The problem posed by the last two examples is this: while applying E(0) to derive the verbal past participle will permit movement to the subject position because it thematically "frees up" the subject position, it does not force movement to take place, but it is only when movement takes place that we get grammatical verbal passives. The problem is not so bad when one notices that movement (of any kind) is not an essential feature of passive in any case; we have the following passives in which no movement has taken place: (30) (a) Es wurde getanzt. "It was danced" (b) ete discute de beaucoup de choses. "It was discussed about many things". (c) It was [reasoned that Bill had left.Jyp

In all of these cases we have a passive with no movement and a dummy subject. What they have in common is that there is no NP to move. Recent proposals in the theory of case suggest that movement is forced to take place when an NP is in a non-case-marking position - the NP must move to a case-marking position in order to get case. Thus if the verbal passive participle does not assign case, direct objects NPs will be forced to move. We may, therefore, view the verbal passive s having two significant features: it is the Output of E(0), which leaves it with no external argument (thus permitting movement to subject position) and it does not assign case, thus forcing movement, when there is an NP object. In the cases just given, we see that the subject position can be freed up thematically (by E(0) ) and no movement takes place, since there is no NP object. Thus these two features of passive are really independent of one another. The examples (30a) and (30c) require some further discussion. The (30a) case shows that English and German are decidedly different: the English translation is ungrammatical, and we have no explanation for this difference. We might suppose that the difference is simply that the English passive participle can be added only to verbs that are transitive - that is,

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

96

verbs which have either an NP or S complement. When it has an NP complement, movement will always be forced, for the reasons just discussed (here and elsewhere); when it has an S complement, movement is not forced, s in (30c). The example (30c) cannot be analyzed s the result of applying extraposition to a structure to which NP movement has applied, since the intermediate structure is not grammatical: (31) *That Bill had left was reasoned.

NP movement cannot take place in such examples, because the verb reason (like certain other verbs, such s hint and/ee/) do not take NP objects, and the taking of NP objects is prerequisite for this construction: (32) *We reasoned it (*That Bill left was reasoned). *We hinted it (*That Bill left was hinted). We hinted at it (That Bill left was hinted at). We believed it (That Bill left was believed).

See Williams (1980) and Koster (1978) for further discussion. We may summarize the difference between the two passives in the following way: The adjectival passive participle is formed from verbs by E(Th), and the output from is an adjective. The verbal past participle is formed from verbs by E(0), and the output form is a verb with the same kind of argument structure of the seems type verb. Thus the two forms differ lexically. They also differ syntactically the adjectival past participle has simply the distribution of an adjective, and there is nothing more to say about it. The verbal past participle has the further property that it fails to assign case, and this has predictable consequences for the kinds of syntactic environments in which it is found. The difference between these two passives can be illustrated by a rather nice comparison with the verb believe (from Williams (1979) ). We find both adjectival and verbal passive of the verb believe. The passive of the verb believe that is "raising" can only be verbal, because the raised NP bears no thematic relation to believe (much less a constant one): (33) John believes there to have been a riot. Therej is believed [t{ to have been a riot.]

Since it must be verbal, due to the failure of thematic constancy, the following are predicted: (34) *How widely believed is John to have left.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

97

Here, the passive is not an instance of E(Th), so the participle must be verbal, but if it is verbal and not adjectival, it cannot be fronted by WH movement. On the other band, there are passives of believe which are thematically constant, and therefore these can be, and in fact are, adjectival: (35) That story is widely believed (story is the theme ei believed). It is widely believed that Bill was here (that Bill was here is the theme of believed)

If these are adjectival, then we predict the grammaticality of: (36) How widely believed is that story. How widely believed is it that Bill was here.

These contrasts ftirther confirm the correlation of E(Th) with the adjectival past participle and of E(0) and NP movement with the verbal past participle. We might consider the question of whether this correlation isaccidental. These facts with believe are rather subtle (they have never been reported in the literature, to my knowledge) and one would thus hope that knowledge of them would not be required to settle questions of English morphology. To see that this correlation is not accidental, consider what would happen if we reversed the correlation - imagine E', just like E(nglish) except that in E' the thematically constant passive is verbal, and the thematically inconstant passive (the Output of E(0) ) is adjectival. Is E' grossly different from E, or is it subtly different from E? Certain considerations lead one to suppose that it might be grossly different from E. Suppose that adjectives do not have NP positions in them but only PP positions, and suppose further that every P is a case assigner. These are certainly reasonable assumptions. It would follow that AP would have no bindable positions, if NP movement trace must occupy caseless positions. Thus the adjectival passive could involve no NP movement, except out of embedded Ss. Thus we would have adjectival passives like the following: (37) Itwashit A of John. Joluij was believedA t{ to have left.

The last possibility exists in E, s well s E', with such adjectives s certain, s in: (38) John is certainA t{ to have left.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

98

E' will not have, however, such passives s: (39) John was hitA (by Bill).

Furthermore, there would be no verbal passives, because it is impossible to have anything external in the category V except Actor, when Actor is a member of the argument structure. This principle was needed, it will be recalled, to explain why Actor is always external in active verbs (and to permit such cases s seem (where nothing is external, and there is no Actor). Thus, the passives in E' would be quite an odd assortment - there would be practically none of the normal passives that are found in E. E', then, even if it is an allowable language, which it is according to our theory, would be extromely different in gross surface form from E and easily distinghuished from E. Given the government theory, we predict further differences between E and E'. According to the government theory (see, for example, Chomsky (1981) and Rouveret and Vergnaud (1980), empty nodes must be "governed"; that is, "minimally c-commanded " by lexical nodes. Given that the English passives are of two kinds, an adjective with one external argument, and a V with no external argument (and NP movement), we predict the following surface structures: (40) (a) (b) John is [knowny t^ have left]yp It is [known^]^y> that John left.

In (a) the complement is inside the VP, s it must be, in order for the trace to be governed by the verb; in (b), the complement is outside of the AP, s it must be in order to be the external argument of the head of the AP. Given this difference in surface structure, which can be inferred from a combination of universal principles and rather gross properties of E, we predict the following differential wrt to the action of VP deletion: (41) (a) (b) *John is known to have left, but he isn't to have gone to the movies. It is known that John left, but it isn't that he went to the movies.

In the (a) sentence, only a subpart of a VP has been deleted, whereas in the (b) sentence, a whole phrase, the AP, has been deleted. As was shown in Williams (1977) it is ungrammatical to delete a subpart of a VP. Again, it seems that a rather obscure fact follows from a combination of universal principles and gross properties of the English passive.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

99

3.2.2 () - "Internalize the external argument": The rules E(X) kept the number of arguments in the input and the output the same. The rules I(X), on the other hand, add one argument to the input argument structure. Since these two rules exhaust the possibilities, we predict that no rule of morphology can shorten argument structure. In Section 4 we will defend this. I(X) is for causativization and for nominalization. Although more complex than E(X) to specify, it admits of no more descriptive freedom. Its action on an argument structure is s follows:

(42)

I(X):
(a) Set the external argument of the input word "equal to" X in the output word; (b) Add a new external argument, A for verbs, R for nouns.

As an example, consider the rule which adds -ize to words. This is an I(X) rule - in particular, I(Th). Suppose we apply this nile to random, which has argument structure (Th): (43) I(Th): randomA> randomy (A, Th=Th)

The argument structure for randomize says that it has a new external argument, A, and that its theme, also new, "corresponds" to the theme of the input word. This last specification serves to capture the fact that the theme of randomize is the thing that becomes random. The rule which adds causative en is also I(Th), and derives words with argument structures just like -ize. Also, the rule which relates intransitive melt to transitive melt (The ice melted; John melted the ice) if this is a rule of morphology, is another instance of I(Th), and derives transitives in the same way that -ize does. We may now turn to a brief sketch of the Japanese causative System. In Japanese, causative verbs are formed lexically from simple verbs by adding the suffix -sase onto simple verbs. We shall see that the resulting verb has an argument structure determined by applying I(Th) or I(G) to the original verb. (The Situation is slightly more complicated in fact - in certain circumstances I(G) forms permissives, rather than causatives, but for our present purposes, this distinction is irrelevant.) As a background, let us assume that Japanese has the following realization rules: (44) Theme: Goal: PP0 PPni

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

100
That is, Themes are located in "o" postpositional phrases, and Goals in "" postpositional phrases. When -sase is added to an intransitive verb, either I(Th) or I(G) may apply (with the difference in meaning noted above): (45) I(Th): V(A) -> Vsase (A, Th=A)

If I(Th) applies, we will get sentences like: (46) John ga Bill o Vsasemasu. "John made BU1 V"

If I(G) applies, we will get sentences like: (47) John ga Bill ni Vsasemasu . "John let Bill V"

When -sase is added to a transitive verb, that is, a verb which takes an "o" postpositional phrase, then only I(G) can apply, not I(Th). There are two ways we could account for this. First, we might point to the fact that the result of applying I(Th) to an argument structure which has a Theme already would result in two Themes: (48) I(Th): (A, Th) - (A, Th, Th=A)

Now, we may ask, is "Th" the same thematic relation s "Th=A"? We have been quiet on this point thus far. Suppose we say yes then this will not be a well-formed argument structure, since we have two instances of the same thematic relation, and since the items in an argument structure are simply labels, we cannot have distinct occurrences of the same labels. But in fact, it appears that causativization by I(G) of a verb which has a Goal phrase already is permitted, despite the fact that this results in two Goal phrases. So perhaps after all, we would like to consider "X" and "X=Y" distinct and prohibit the application of I(Th) to transitives by a restriction on the number of "o" phrases that may appear in a surface structure clause. This latter type of explanation is similar to that set forth in Aissen (1974), but without her notion of "clause union". Nominalizations and causatives may seem to be an odd natural class, but it seems that nominalizations are also instances of I(X) rules. When a verb or an adjective is nominalized, its external argument is made an internal argument of the noun. We will examine two cases of this in English, I(A)andI(Th).

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

101

In the formation of -ing nominals, either I(Th) or I(G) may apply. Recall that the (at least unmarked) external argument for nouns is R: (49) I(Th): growly (A) -> growlingN (R, Th=A) "the growling of the lions" I(A): growly (A) -+ growlingj^ (R, A=A) "the growling by the lions"

As in the case of the Japanese causative, if we have a transitive verb with theme object, then we cannot apply I(Th), but only I(A): (50) the shooting of the lionsj by the hunters l l *of the hunters J

This example perhaps suggests the answer to the question raised earlier about the Japanese causative. If we assume that shooting is always transitive, then we can derive the two argument structures: (51) (A, Th)^ (R, Th, Th=A) ( , Th) (R, Th, A=A) (by I(Tli) ) (by I(A) )

In fact both of these can be found for the nominalization ofshoot (shooting\. (52) the shooting of the hunters the shooting of the lions by the hunters

U seems that we do want to prohibit the structure (R, Th, Th=A), we simply want to prevent two o/phrases, each linked to a role of theme, to appear in surface structure. The suffix -ion works in much the same way: either I(A) or I(Th) may apply: (53) I(Th): the pontification of the Pope I(Th): the pontification by the Pope (from pontificate (A) ) (R, Th=A) (K, A=A)

One prediction unfortunately made by the System thus far represented is that intransitive verbs with external themes (like the intransitive melf) should be able to undergo I(A), which they cannot: (54) I(A): melt (Jh) -> melting (R, A=Th) "the melting by the snow"

I have no particular insight s to why this should be so, but I do have a

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

102

couple of suggestions. Suppose that we have a hierarchy of thematic relations, s proposed in Jackendoff (1972), which runs: Agent, Theme, Goal, ... and suppose that we stipulate that in the "X=Y" Statements derived by I(X) that X cannot be higher than Z on the hierarchy. This will prohibit the above derivation of melting, but it will permit all other derivations that we have considered. We might take a somewhat different tack. Suppose that the normal case for nominalizations is 1(0), where this is interpreted to be: (55) 1(0): (a) add a new external argument (the appropriate one for the category created; R for nominals, etc.)

Since the new argument is external, the external argument of the input is made internal, by automatic convention (s in the case of E(X)). A problem could arise if the new external argument was the same s one of the old arguments, since we would then wind up with identical arguments. This could not arise for deverbal nouns, since the external argument added to make a noun is R, and R is never a verbal argument. But the problem would arise for denominal nouns (e.g., tableness). Let us assume that the normal rule for nominalizations is 1(0) and that the problems mentioned with it above are solved. Then, the expected cases wfllbe: (56) the melting of the snow the shooting by the hunters

since in these cases the argument types are preserved: snow is the Theme of melting, just s it was the Theme of melt; hunters is the Actor of shooting just s it was the Actor ofshoot. From this point of view, one of course would not expect to find such cases s "the melting by the snow" since in such cases the Theme of the input (melt) has become the Actor of the Output (melting). That is a good result, but one would also not expect the form, the shooting of the hunters, which is well-formed, since here s well the thematic roles differ in the input and Output. For such cases one could either apply I(Th) s a special "marked" alternative, or, one could say that PPof was a special "marked" realization of Actor in NP. I will leave the problem s it Stands. 1(0) s the unmarked case for nominalizations does well by adjectives, though, if Theme is the external argument for adjectives. Consider the suffix -ness, for example, s an example of 1(0) (or perhaps I(Th) ): (57) 1(0): complete (Th) *> completeness (R, Th) The completeness of the results.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

103
3.3. Concluding Remarks To summarize the point of this section: once you divorce the rules which determine the syntactic realizations of various argument types from the rules which relate the argument structures of morphologically related words, then this latter class of rules can be stated in a quite restrictive way. What these rules seem to boil down to is that you can affect the external argument of the input word in one of two ways: you can make one of the internal arguments into the external argument, with a consequent demoting of the old external argument to internal Status (E(X) ), or you can add on a new external argument, again with consequent demoting of the old external argument to internal Status (I(X) ). It is important to emphasize that grammatical relations play no role in this System, or, to the extent that they do, there are only two: subject and nonsubject (where I use internal and external instead). In this System outlined here, thematic roles are tied directly to syntactic realizations, by the realization rules. In a theory with grammatical relations, this relation is mediated by the set of entities called grammatical relations, and there are two sets of realization rules: one set which says which thematic roles are realized s which grammatical relations and another which says which grammatical relations are realized in which syntactic positions. Since there are no grammatical relations in the System presented here, there can of course be n o rules, syntactic or lexical, which operate on representations which contain or are composed in terms of grammatical relations. Therein lies part of the restrictiveness of the System presented here. Another portion of the restrictiveness of the theory presented here lies in the Substantive claim that morphological rules can affect argument structure only by affecting the external argument (I(X) or E(X) ). This is far from obviously true; in the next section, we will defend this view from one kind of attack on it, the case of lexical "detransitivization" rules. In fact, any putative lexical rule which does not affect the external argument of a lexical item cannot be a lexical rule, according to the theory presented in this paper. Proposals for such rules are quite common. A good example is the rule of dative movement in English, which, it is argued by some, cannot be a transformation - and thus must be a lexical rule (see Oehrle (1976) ). We may agree with Oehrle that this rule is not a transformational rule, but we need not conclude that it is a lexical rule - in fact there is no need for it to be a rule at all. First of all, it cannot be a lexical rule, because it cannot be an instance of either E(X) or I(X), since the external argument of the input (no matter which way the rule runs) is not affected. Second, it need not be a lexical rule, if we consider the foUowing line of thought: Suppose English has two ways of realizing Goal, s outlined in Section 3.2.1 s NP2 and s PPto. Any verb with a Goal

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

104

internal argument must subcategorize for one of these two possibilities, or for both. In the case of verbs that use both realizations, there is no need to say anything other than that they use both there is no need, in particular, to consider such verbs to be two verbs and to relate these two verbs by lexicalrule. It is telling that there is an affix (-en) for passive, but no affix for dative movement. Why is this? According to the theory presented here, there could be no such affix s a "dative movement" affix, because there can be no such lexical rule, and this is because the relation that would hold between the argument structures of the forms related by the rule could not be stated (is neither I(X) nor E(X) ). A rather strong (embarrassingly, I am sure) prediction is made: languages can have all sorts of passives, with affixes - for example, a passive which moves locatives and themes and adds a different affix for each of these and these would be cases of E(X). But no language could have a lexical rule like English dative movement, nor, of course, an affix associated with such a rule, since such rules do not affect external arguments.
4. LEXICAL DETRANSITIVIZATION VS OPTIONAL SUBCATEGORIZATION

There are two ways one might represent the relation between transitive eat and intransitive eat. One might use the notation of optional subcategorization and write "eat, (NP)." Or one might have two eats one transitive and one intransitive, related by a lexical rule of detransitivization: (58) V[+obj] *V[-obj]

For simple cases, it is difficult to distinguish these empirically. But for some more complicated cases, a difference in prediction can be derived. The transitive-intransitive relation is sporadic not all transitives have corresponding intransitives. Thus put and attempt do not, while eat and hope do. In the subcategorization theory, this means that some frames will have parentheses in them, and some will not. In the lexical detransitivization theory, this means that the rule of detransitivization is not completely general. This consideration is certainly not decisive, since lexical rules are in general not completely general. Put is a case where two things are obligatorily present, an NP and a PP. Position, on the other hand, whfle it has two arguments in the VP, has only one obligatory argument. And promise, which has two arguments s well, has both of them optional. The subcategorization theory predicts that four different cases will be found for verbs with two arguments in the VP:

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

105

(59)

XY one expansion: X () two expansions: X , () two expansions: , ()()four expansions: , , , 0

put make (see below) position promise

All of these are attested, except for the second. As a case of the second, we might consider the dative that occurs in such cases s: (60) Make Bill a cake.

This dative is always optional and precedes a direct object. Now the subcategorization theory predicts that a verb with two arguments in the VP will exhibit one of these four "expansion sets" but no other. There are other expansion sets, however, besides the ones predicted to occur by the subcategorization theory. For example, the expansion sets

(61)

X Y,X,0

X,Y,0

are conceivable but are predicted not to exist by the subcategorization theory. In the case of the second, for example, in order for 0 to belong to the expansion set, both X and must be made optional; but if both X and are optional, this implies that will belong to the expansion set. The lexical detransitivization rule, however, makes no such prediction. Consider the hypothetical lexical item M with two arguments in the VP, X and Y. First we derive M' by applying lexical detransitivization to eliminate the argument ( s we might to derive "position NP" from "position NP PP"). Then, we apply lexical detransitivization again, to derive M", just like M', except that now the X argument is removed:

(62)

>M'

>M" 0

We see that this triple (M, M', M") exhibits an expansion set which we saw was predicted not to occur by the subcategorization theory. Ingeneral, the lexical detransitivization theory predicts a wider r nge of possibilities than the subcategorization theory. To the extent that exactly the r nge of cases predicted by the subcategorization theory is found, and no others, that theory is to be preferred. Note that it will not do to object that the diagram above is "incomplete" in the sence that lexical detransitivization has not applied to M to derive M'", which has only s an argument. This objection is pointless because we know from the Start that the transitive-intransitive relation is sporadic, and the only real question Brought tois the r| Chinese University of Hong Kong (Chinese Univers is what you by nge of possibilities.
Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

106

More is at stake than whether the transitive-intransitive relation should be done by one mechanism or another. The theory of lexical rule is at stake. I have proposed in this paper a theory of lexical rule that will not countenance such a rule s lexical detransitivJzation, for the simple reason that such a rule does not affect the external argument of words (i.e., it is not an instance of either I(X) or E(X) ). Such a role would affect only the arguments internal to a VP. Thus it is encouraging to find that such a rule has other faults s well when compared with reasonable alternatives. To summarize, the rule of lexical detransitivization has two significant faults: it predicts that a broader rnge of "expansion sets" will be attested than the subcategorization theory, and it forces the abandonment of anattractively restrictive theory of lexical rule.
5. THE FRENCH MIDDLE CONSTRUCTION

The theory of morphology outlined here, in conjunction with the theory of predication in Williams (1980) affords some explanation of aspects of the French middle construction, with some help from recent ideas of case theory. The French middle is a construction in which (a) the normal object appears in subject position, (b) the verb is in the active form, and agrees with the subject; and (c) the reflexive pronoun se is attached to the verb. The following is an example: (63) Ce livre se vend This book sells itself

We will proceed with our discussion of the middle, first by assuming a certain representation for the middle, and then considering why it has to have that representation. Suppose that there is a lexical rule for forming middle verbs. The rule does not alter the verb morphologically; it does, however, alter a verb 's argument structure by the function E(0). Now suppose that we have a verb like vend with argument structure (A, Th). When we apply this rule, we will get vend (A, Th). Next, suppose that we generate this verb with a direct object reflexive clitic and a subject, s follows:

(64)
0

The "0" under vend is simply to indicate that the verb is a 0-external argument verb. When the rule of clitic movement and the rule of predication (see Williams (1980) ) have applied to this form, we will derive:

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

107

(65)

NPitsej
0

This representation seems to capture the basic facts about the middle construction. The surface subject is interpreted s the direct object (here, it winds up coindexed with it) and the verb bears a reflexive clitic. Furthermore, the superficial subject has the thematic relation of the object, since it is coindexed with it, and the subject position itself receives no thematic relation, since the verb form was derived by E(0), and thus has no external arguments. Given the deep structure described above, we can derive a reasonable account of the middle. But why this deep structure? In other words, why can't the middle verb (derived by E(0) ) appear in other kinds of deep structures? If we can answer this question, then we can be said to have explained some properties of the middle construction. First, why is a reflexive clitic involved? And why does it wind up coindexed with both the trace and the NP subject? To answer the second question first : the reflexive is coindexed with the trace by movement - all cHtics must move, by definition; the clitic must be coindexed with the W that contains it by the requirement stated in Williams (1980) (s a variant of the SSC) that a reflexive cannot be free in a VP (and where being coindexed with the VP counts s being bound within the VP); the VP and the subject are coindexed by the rule of predication; therefore, by the transitivity of coindexing, the subject wfll be coindexed with the clitic and the trace of clitic movement. Now for the first question: why is a reflexive clitic involved? This question divides into two parts. First, why can't there be a nonreflexive clitic, or ordinary lexical NP, and second, why can't there be no clitic i.e., why can't NP movement move a direct object of a middle verb directly into subject position. The reason a nonreflexive clitic cannot be used is this: again, a nonreflexive clitic must move (since it is a clitic); furthermore, a nonreflexive clitic must not be coindexed with the VP that immediately contains it (again, a variant of the SSC, s given in Williams (1980)); hence, we will get the structure:

(66)
0

Here, the subject NP is not coindexed with the object trace. Since it is not coindexed with anything that is an argument of vend, and since it cannot be interpreted s the external argument of vend, since that verb is in the middle and thus has no external argument, then the subject NP

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

108

can receive no Interpretation with respect to vend, and this is why sentence is bad. Thus no nonreflexive clitic can be used. A similar planation can be given of why a fll NP cannot be used in place of reflexive clitic. Now, why must there be any clitic at all? Why not derive it by movement, from subject to object position, giving the representation:

the exthe NP

(67)

N 0

Here, the NP subject will receive a thematic Interpretation, by virtue of bin ding the trace, which occupies an internal argument position. So the theory of argument structure is not really going to explain this. But this is where the theory of case Steps in. By current theorizing on case (see, for example, Chomsky (1981) ), case must not be assigned to the trace of NP movement. If the active form of the verb assigns case, then the trace above will receive case, thus ruling out this example. The past participle of the passive construction, on the other hand, does not assign case, and so NP movement is permitted in the derivation of passives the trace left behind will not get case in that construction. A bit more needs to be said, because the stipulation that traces cannot get case, or be in case-marking positions, will also rule out the actual middle construction, since it has a trace in it. The crucial property of the trace in the middle construction is that it is a clitic trace and not an NP trace. Clearly, we want to say that the clitic trace is exempt from the case-marking restriction that the NP movement trace is subject to. We might do this in a number of ways, but, however it is done, the solution will not apply simply to the middle, but to all uses of clitics in all constructions, and so the solution will not be an ad hoc answer to our current concern; that is, the same question arises for the use of non-reflexive clitics in non-middle constructions:

(68)

Jean [le{ mange t{]

One way to do this is to say that clitic movement takes place after case marking. The exact mechanism does not concern us here, though, once we have noted that the problem immediately generalizes to all clitics in all uses. From two assumptions - that middle verbs are formed by E(0) and that active verbs are case assigners - we have explained a number of facts about the middle construction: why it has a clitic, why that clitic is reflexive, why the subject is interpreted s the object argument.2 I regard this s a fair attempt at explanation, but there are some problems, and

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

109

some further speculations. Some of these arise with the English middle construction. English has two middle constructions. One of them is just like that in French, except of course English does not have clitics. Suppose that there is a middle formation rule in English s well, and that this rule is an E(0) rule. Then in English, we can derive middles like: (69) This book sells itself

which will have the representation: (70) This boolq [sells itselfj yp ! 0

This is exactly like the French middle, and its properties can be explained in exactly the same way s the French middle: the reason a reflexive is involved is that otherwise, the subject NP could not be coindexed with anything in the W, and if it were not, it could not receive any thematic Interpretation (the verb being a/) external argument verb).3 This is an interesting construction from the point of view of the "theta criterion". In this construction, an NP gets a thematic relation by virtue of being coindexed with a reflexive pronoun which occupies an internal argument position. In the more ordinary use of reflexives, the reflexive and its antecedent are independently assigned thematic roles, and it is usually thought that it is only through mvement that an NP will get its thematic role by virtue of what it is coindexed with. These cases refute such a view. Furthermore, the facts of such cases fall out automatically in fact the same s in French, modulo the clitic/nonclitic distinction which divides these languages independently. One concludes from such examples that the faflure of case assignment to object position does not always attend the application of E(0). These are independent. Problems arise when we consider another English middle construction, which differs from the English reflexive middle construction and the French middle construction in not having a direct object reflexive: (71) These books seil well.

There is no problem in saying that these cases are the result of applying E(0) to verbs; the problem is that we will have to use NP mvement to derive the appropriate surface forms: (72) These books. [seil t{ well.] yp 0 '

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

110

The problem with this representation has to do with case assignment: if active verbs must assign case, then we have no explanation of why this form is permitted to have a NP movement trace after the active verb. We might suppose that there is a special rule which removes the case-assigning property from these verbs, despite the fact that they remain active verbs. This would weaken our explanation of why the French middle used a reflexive clitic instead of NP movement, but in fact we now need some way to distinguish this middle from both the French middle and the English reflexive middle, and case assignment could be the way to do that. In sum, then, the description of the middle involves two different theories. First, there is the theory of argument structure. As far s that theory goes, the middles are all E(0). Second, there is the theory of case. If the middle verb is a case assigner, then we will get a reflexive object; if not, NP movement. In either of the latter two cases, the surface subject gets its thematic relation by virtue of being coindexed with something inside of the VP. The one further difference between the French middle and the English reflexive middle namely, that the reflexive moves in French, but not in English follows from the theory of clitics, and the fact that French has them, but English does not. One virtue of the analysis presented here is that the reflexive clitic in the middle construction has the properties of any clitic in any other construction, and there is no need to replicate the basic property of clitics ("they move in front of verbs") inside the theory of the middle. Nor does the reflexiveness of the reflexives in the two constructions require any special treatment: in the case of the middle, s in the case of any other use of the reflexive, we need only say that bound anaphors are subject to certain opacity conditions. Thus a good deal of the description of the middle reduces to nothing special: the theory of bound anaphora and the theory of clitics. In not every imaginable treatment of the middle, however, is this reduction possible.
6. SUBCATEGOR1ZATION

We have ignored the question of subcategorization so far, except to rely on it for explanations that were not supplied by other parts of the theory. We have assumed more or less theAspects theory of subcategorization, where a head of a phrase can determine the category of the phrases that appear in that phrase with that head. We will now look more closely at those proposals about subcategorization which will be consonant with the theorizing we have done thus far. Of course, the application of a morphological rule changes subcategorization, but the question is, do we need a set of rules for specifying the changes that can be effected in subcategorization? The best answer is no -

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

111

that any changes in subcategorization will follow from other changes induced by the application of a rule, most likely the changes in argument structure and in case-assigning properties. In this section, we will strive for this best answer; in particular, we will outline a theory by which the argument structure and the category of the derived word will determine the subcategorization of the word, thus avoiding the need to specify the subcategorization of the Output of a morphological rule. The theory assumes a "markedness" analysis, or ranking, of the realization rules. As can be seen by inspecting (18), there is potentially more than one realization rule for each argument type. Thus, for example, Goal can be realized s a NP2 or s the NP in a PPto. If a particular verb has a Goal argument, it must subcategorize for one or the other or both of these options. Now suppose that we say that the realizations for any particular argument type are ranked, and that for Goal in English, PPto is first, and NP2 is second. One use we might make of this is to predict that if a new monomorphemic verb were invented, and that verb had a goal, then it would have the PPto subcategorization, and not the NP2 subcategorization. We might also expect that a new morphologically complex word would have an unmarked realization of Goal, if it had a goal; thus, we might automatically supply new words with unmarked realization possibilities, regardless of the source. Other factors can prevent the manifestation of the unmarked case. For example, of the two ways to realize Theme, s NP or s the NP of PPof (s in think of) surely the bare NP is the unmarked case. But in the complement structure of NP, bare NP is not allowed at all, presumably because case cannot be assigned to it there (N does not assign case) and so we find only the marked realization of NP. There remain several further questions to which we can give only tentative answers. For example, we know that subcategorization is sometimes obligatory ( NP) and sometimes optional ( (NP) ). Do morphological rules affect the obligatory/optional Status of Output words? The suffix -ize, for example, seems to create obligatorily transitive verbs: one says, John rubberized the raft, but not *John rubberized. Again, we might suppose that " NP" is the unmarked choice for internal theme, and " (NP)" is the marked choice. A further question is, what Information is subcategorization sensitive to? For example, is it sensitive to case? At first glance it appears to be; in German, for example, certain arguments are sometimes instantiated s genetive NPs for certain verbs, and this is idiosyncratic; presumably, verbs must be subcategorized for such cases. And there is some indication that in English, subcategorization must refer to heads of PPs. A perhaps crucial pair of cases is give and present, both of which have goals, and both of Brought to you by which have " NP PP" subcategorization: | Chinese University of Hong Kong (Chinese Univers
Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

112

(73)

I presented it to him. I presented him with it. I gave it to him.

But, if all we say about give and present is that, we predict: (74) *I gave him with it.

There are two Solutions: we can either distinguish "NP" and (which we earlier called "NP2") or we can distinguish PPto and I In the former case, we will give the subcategorizations in (a) below, and in the latter, the ones in (b): (75) a. Present: _NP2 PP Give: NPPP b.Present: _NPPPwith Give: _NPPPto _NPPP _NP 2 NP _NPPPto _NP 2 NP

It is unclear whether such Solutions will be available generally, or whether they should be allowed generally. A possible objection to the above Solutions, which allows the subcategorization restrictions to "look" at the heads of phrases, is that this is contrary to the claim made earlier in the paper "that words do not have realization schemes, languages do" since by careful subcategorization a verb can determine uniquely which realizations it will be subject to. While the ability to subcategorize for heads of complements may in fact allow a verb to uniquely determine which realizations it will be subject to, this does not reduce the original claim to vacuity, since there will still only be a few fixed realizations for any argument type, and only these will be available. Thus, no English verb, for example, can determine that its Goal wfll be realized s PPfromWe have a theory now where subcategorization can refer not only to phrases, but also to heads of phrases. Perhaps we could make the theory at least sound more restrictive by allowing subcategorization for heads only, but heads under different descriptions. Thus, instead of PP, we have P, and by the same token, instead of PPto we have 70, and subcategorizations for -bar s well (or to-double-bar). And S could be referred to s tense-bar, if tense is the head of S.
l. (i) NOTES Even though Bill is not the subject (or external argument) ofseen in this example, Billj was [seen tj]

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

113
(rather, it binds the intemal argument position occupied by trace), we still would like to say that Bill is the subject of the VP "seen tj" in the sense of Williams (1980). Bill is the subject of "seen t" not by virtue of being an argument of the head of "seen t" (i.e. of seen), but rather by virtue of binding the trace. This the predication involved in the verbal passive is not of the type called "simple headed" in Williams (1980) but rather of what was called there "complex," since the predicate "seen t" is a predicate only because it contains a syntactic variable (t). The fll representation of this sentence in the notation of Williams (1980) would thus be (ii) Billj was [seen tj] vp.

Raising would work similarly; although Bill is not the subject of seems, since, s mentioned before, seems has no subject (or external argument), still, Bill is the subject of "seems t to be sick," and using the notation of Williams (1980) we have: (iii) Billj [seems [tj to be sick] ] yp.

Again, the VP "seems t to be sick" is a complex predicate, that is, a phrase that is a predicate by virtue of containing a syntactic variable. 2. It is worth notin g that the only difference between the middle and the ordinary use of the reflexive is that in the middle, the verb is a 0-external argument form, but in the ordinary use, the verb is an ordinary l-external argument form: (i) Middle: Le livre{ [sq vend tj] 0 [se^ lave tj]
4

The book sells itself"

"Ordinary": Jean^

"John washes himself"

In the middle, there is only one thematic position, the object position, and the subject gets its thematic role by virtue of being coindexed with the trace in object Position. But in the "ordinary" case, there are two thematic positions, the subject and the object position, and each receives a different thematic role. 3. Many such cases are listed in Fiengo (1974). The construction is also quite common in Dutch, formed on zichzelf (Riny Huybregts, personal comm.). REFERENCES

Aissen, J. 1974. "Verb Raising." Linguistic Inquiry, 5. 325-366. Allen, M. 1978. "Morphological Investigations." University of Connecticut dissertation. Aronoff, M. 1972. "Studies in Analogical Pseudosyntax," no. 1. MS. Aronoff, M. 1976. Word Formation in Generative Grammar. MIT Press: Cambridge, MA. Bresnan, J. 1972. "The Theory of Complementation in English Syntax." MIT dissertation. Bresnan, J. 1978. "A Realistic Transformational Grammar." In Bresnan, Halle, and Miller, eds., Linguistic Theory and Psychological Reality. MIT Press: Cambridge, MA. Chomsky, N. l965.Aspects ofthe Theory of Syntax. MIT Press: Cambridge, MA. Chomsky, N. 1980. "On Binding." Linguistic Inquiry, 11. l Chomsky, N. 1981. Lectures on Government and Binding. Foris Publications: Dordrecht. Fiengo, R. 1974. "Semantic Constraints on Surface Structure." MIT dissertation. Gruber, G. 1976. Lexical Structure in Syntax and Semantics. North Holland: AmBrought to you by | Chinese University of Hong Kong (Chinese Univers sterdam.
Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

114
Jackendoff, R. 1972. Semantic Interpretation in Generative Grammar. MIT Press: Cambridge, MA. Kayne, R. 1975. French Syntax. MIT Press: Cambridge, MA. Koster, J. 1978. "Why Subject Sentences Don't Exist." In S.J. Keyser, ed., Recent Transformational Studies in European Languages, MIT Press: Cambridge, M A. Oehrle, R.T. 1976. "The Grammatical Status of the English Dative Alternation." MIT dissertation. Rouveret, A. and J.-R. Vergnaud. 1980. "On Specifying Reference to Subject." Linguistic Inquiry, 11.1. Wasow, T. 1977. "Transformations and the Lexicon." In Culicover, P., T. Wasow,and A. Akmajian, eds., Formal Syntax. Academic Press: New York. Wasow, T. 1978. "Remarks on Processing, Constraints, and the Lexicon," In ProceedingsofTINLAP2. Williams, E. 1977. "Discourse and Logical Form." Linguistic Inquiry, 8. 3. Williams, E. 1979. "Passive." MS. Williams, E. 1980. "Predication." Linguistic Inquiry, 11.1. Williams, E. 1981. "On the Notions 'Lexically Related' and 'Head of a WordY' Linguistic Inquiry, 12.2.

Max Planck Institute for Psycholinguistics Berg en Dalseweg 79 Nijmegen, Holland and Dept. of Linguistics South College University of Massachusetts Amherst, MA 01003 U.SA.

Brought to you by | Chinese University of Hong Kong (Chinese Univers Authenticated | 172.16.1.226 Download Date | 3/4/12 2:05 PM

You might also like