Next >>

Regardless of the convoluted pondering over covert/overt operations concerning the Subject here (e.g., Agr_s, Spec-TP, mult-Spec, etc.), the overall structure nonetheless must project a partially-fledged IP (via TP). In other words, since the T component of IP (in the Pollockian sense) has already been established at this stage-2, the reduced VP-stage (as seen in stage-1) can't suffice. This is the crucial difference between my stage-1 (where no IPs were reported at all) and stage-2 (where the acquisition (albeit not the mastery) of IP was spotted. In sum, this has the flavor of saying that although all feature operations may have their initial locus at LF, only in the sense that procrastinate seems to prefer covert operations to overt ones, there is an added stipulation which states that once an IP projects overtly at PF (creating a functional checking domain), all subsequent clauses thereafter must theoretically project an IP (albeit minimally via TP). Again, if Agr were to have no strong features at LF, PF considerations would give no reason for it to be present at all (Chomsky op.cit: 351). Therefore, the stipulation of T suffices to force us into projecting an IP for (10 e-j), even though the relevant features/categories of the clause seemingly project a VP. The distinction of duty between overt and covert movements may be expressed by the fact that overt movements need to carry along whole categories for PF convergence (i.e., there is no sense describing a covert PF as there is no feature strength distinction having to do with phonological features), while in covert movement features raise alone. Hence, once an XP has been acquired (at PF) by the onset of the category <x> being carried, the established XP is in place and avoids vacuous projections at all expense.

(The Roles of Agr vs. T )         The roles between Agr and T can be further reduced. Recall that the majority of early Nominative Case usages found were among ambiguously marked finite/nonfinite structures. Among such early constructs, a minus-Tense/nonfinite interpretation makes the most sense. There are two clear reasons for this. Firstly, at the onset of verbal 3sg Nom Case constructions, the present tense marker "s" is left out (as in 10). (See §1.2.3 for Tense). Secondly, in Double-Verb (SVV) constructions, e.g., Infinitive constructions with a Nom Case, the infinitive particle "to"--which arguably has anaphoric/infinitival tense and is matrix-bound by INFL--is always missing. (In fact, the infinitive particle "to" doesn't seem to emerge until late in File 23 (3;2); however, when it is used, the matrix clause shows proper tense). An argument can be made that when there is no "to", there must also be no tense in the matrix INFL to bind T in the complement clause. In other words, although Agr(eement) is realized on the Nom subject of these Double-Verb constructions, the main verb's T(ense) is assumed not to project--much in the same manner as the finite particle 'to' is assumed not to project. Consider such early examples of [-T/+Agr] SVV constructions with mostly ambiguous finite main verbs:

(12)                        Nominative SVV [+Agr/-T] with Infinitive "to" omission:

                        a. I want ø kick            (file 8: 2;4)        d. She going ø touch my man    (3;2)

                        b. I want ø cook           (file 14: 2;7)      e. You want ø help me?(file 22: 3;0)

                        c. I want ø hit a spider  (file 22: 3;0)      f. I want ø write            (file 24: 3;4)

(Nb. There are no counter examples of an apparent Tense bound matrix clause without a "to" infinitive complement: e.g., He wanted/wants go home, etc., is not attested in my data. Such examples falsify my argument here that "to" has tense and binds tense to its matrix INFL-clause.)

Let us embark on Schütze and Wexler's (1996), Schütze's (1997) discussion that seeks to analyze (inter alia) the Present Tense "s" suffix as unambiguously signaling the presence of tense and agreement. Such a description would predict Accusative subjects never to occur alongside the suffix "s". E.g., Wexler points out that the combination "Him cries" is unattested. However, in my own data, and in a wide array of literature found elsewhere (e.g., Huxley (1970), Aldridge (1989), among others), such combinations are in fact reported. One interesting way, though, in which we could save Wexler's elaborate paradigm would be to suggest that the English suffix "s" doesn't mutually signal T and AGR, but rather exclusively signals Tense--the suffix "s" would then have no overt bearing on Agreement at all (i.e., AGR simply remains indifferent to the suffix "s" due to the nature of an invisible agreement assigning mechanism in English (e.g., He/She cry [+Agr, -T], Schütze et al. op.cit: 9), though it may coincidentally sit among the presence of the Tense marker suffix "s" for 3sg.[i] Consider the revised suffix "s" paradigm below:

(13)

Suffix "s"=>                 I.          +Marks Tense [ii]   (Radford: lectures '97)

                                                                        a. Him cries (-Agr)

                                                                        b. I works (+Agr invisibly marked)

                                    II.         -Marks Agreement: (+Agr invisibly marked)

                                                                        c. (He cry)

                                                                        d He crie-s- ([+T], "s" doesn't mark Agr)

                                    III.       +Marks Tense (+Agr invisibly marked)

                                                                        e. He cries

Furthermore, following Radford, if we assume that some children have the following entries for subsequent inflections:

(14)                        a. +d if past tense

                               b. +s if present tense

                               c.  ø  otherwise (perhaps as a universal default)

and, if we assume (contra Schütze: 1997) that T and Agr are not fused together and optionally projected at PF, we would then expect to find the following paradigm of early utterance types (features in brackets are those features carried by the Verb/INFL):

(15)                                    Non-Fused T/Agr Paradigm

                        a. I/He cried =>[+T,+Agr]                    e. Me/Him cried =>[+T,-Agr]

                        b. I cry =>        [-T*, +Agr]                  f. Me/Him cries =>[+T,-Agr] [iii]

                        c. Me cry =>    [-T*, -Agr]                   d. I/He cries =>[+T, +Agr]

(* no indication of Tense--used both in past/present contexts)

Noting in (13) above the phonological sameness of II and III with regards to cries (ex. d, e)--viz., it remains impossible to tell in English whether or not only Agreement projects (as opposed to both agreement and tense) regarding the suffix "s". However, what we gain by postulating example (c) in II (illustrating no potential correlation between the suffix "s" and +AGR) is an added feature in I which now can indicate a potential T(ense) without Agr(eement). In treating the suffix "s" in such a restrictive manner, a feasible account can now be made showing how a T feature ("s") could go omitted, whilst Nominative Case via invisible Agreement is maintained [-T, + Agr]: e.g., He get a bat, He do it, etc. (cf. (10) above). (Here, Agr-features on I are checked by the Nominative Subject, et vice versa). Theoretically speaking, the converse then holds with respect to utterances containing [+T, -Agr]: e.g., Him goes, Him cries, Him is hiding, me walked, me broke, etc[iv]--all violating the D-I correction (cf. 1a)--though Wexler argues against this. (Wexler makes the distinction that only the "s" suffix--and not the "ed" suffix--is associated with person/number features as well as tense.) Nevertheless, we may wish to claim here that while the "s", "ed" specifically mark Tense, they fail to signal any Agreement--prompting rather a Default Agreement marker (Radford: lectures '97).

Similarly, examples of an overgenerated/default suffix "s", may likewise be interpreted in ways which attribute "s" particularly to T only and not Agr (though Agr is correctly marked by the Nom Subject in ex. a-c) (p.c. Radford). Consider the following token examples of the first emergence of "+S " (file 23) found in my Data:[v]

(16) 

(16')

An additional point to make here--giving further empirical support to the above claim--is that we observe "s" as only marking T and not Plural Agreement--i.e., the plural marking "s" on the Noun (book) escapes verbal Agreement (are) (cf.16'k). The suffix "s" in such examples (albeit few in number) might be interpreted as rendering an overgeneralization effect to that described above: specifically, while the suffix "s" doesn't project its proper person/number Agr-features--e.g., rendering instead an impermissible 1sg Nominative (cf. ex. a-c) (in ways similar to e.g., Him cries)--it does however project a sort of default present tense. An overgeneralization of the "s" may be construed in the light that the present tense suffixes on 1,2sg/1,2,3pl main verbs must be represented by a null constituent [ø]. The fact that Nominative Case is assigned nevertheless under such a confused state, I think, goes to the heart of the issue that the suffix "s" plays only an overt unitary role and not a dual role for the child (i.e., it exclusively marks tense)--with the assumption that Agreement may be marked incidentally by an invisible agreement mechanism (in English): (e.g., I/You/She/He/We/They hate syntax).

In sum, it is clear at this stage that the child has acquired the +Interpretable Tense feature of "s", so [+s] is used whenever INFL has [+T present]. (Unlike the adult specification that calls for "s" iff +3Per-Sing-Pres, the child's entry of "s" only refers to the feature present tense and may not initially adhere to person.) But INFL may optionally project Agr features as well. Whether or not INFL projects an Agreement feature [present +Agr] or [present -Agr], [+s] will continue to be used regardless of Agr. The fact that the child may have access to Agr--as signaled by the case of the subject--speaks only to the notion of the Agreement mechanism itself as cited above. Moreover, the earlier observation that certain aspects of Tense may actually be acquired earlier then Agr, I think, reinforces the previous notion put forward that children may first acquire (semantic) +Interpretable features (Tense) of the suffix "s" and only later do they come to acquire its -Interpretable features (Agreement). This Discontinuity between the child-adult grammars may stem from this notion that only +Interpretable features come on-line at the earliest OI-stage. In this sense, the improper "s" in (16') is restricted to Tense and thus doesn't involve itself with the (invisible) Nominative assignment [+ Agr].

Recall at the end of (§1.1), we discussed the possibility that IP may split into two merger operations--merger at PF and merger at LF. The above analyses of "s" as a Tense feature here further adds empirical support to that notion: examples (cf. 10) a. He cut the tree, c. He get a bat, d. He do it all demonstrate an AGR_sp [=assigning Nom case] merging at PF while TP [=Tense operator] is covertly functioning at LF. Again, one crucial advantage for categorizing "s" here as strictly a Tense feature (and not a hybrid of T&Agr) is that we can achieve clear-cut derivations of T and Agr merger operations at either level of PF/LF. As an interesting side-note (pointed out to me by Radford p.c.) these assumptions jibe nicely with what we know about "South-Western British English" which overgeneralizes the suffix "s" throughout the present-verbal paradigm: (cf. I/We/You/She/They hates syntax (Radford 1997ms ch. 10: p7)).

1.2.3 Possessive Structures

The following section is two-prong in nature. (i) Regarding a stage-1, I claim that this initial stage manifests no inflections whatsoever: contra Wexler, I find no clear evidence of Optionality. Following suit, I then compare and contrast the data (Tables 1.4 & 1.5)--serving as a means to illustrate this Non-Inflectional Stage-1 vs. an Optional Inflectional Stage-2. (ii) In acknowledging a Wexlerian Optional-Stage for our second stage, I put aside general issues of Optionality and proceed to shed empirical light on the generally held hypothesis that No Correlation necessarily holds between D and I outside the appropriate checking domain (e.g., there should be no correlation say between a lower Object-DP and INFL).

(Previews)       Stage-1 (see Data section below) suggests a Prefunctional Stage (viz., an early stage which preexists underspecifications). N+N (=Genitive) constructions are used alongside sporadic usage of (non-verbal clause DPs) My/Mine. (The usage of the definite determiner The is fully productive from the earliest files (cf. §1.2.1)). Early usage of the determiners My and Mine may indicate that these Poss(essive) Nom(inals) are, in fact, Analogous to the robust early (file 1-7) usage of D(P) The (=Analogical-The)--i.e., they comprise of Default Case without their formal features of Genitive Case and don't incorporate verbal elements.[vi] Hence, all DPs (alike) at this stage would possibly mark for +Def(initeness) only. In this sense, the young child may freely alternate between Analogical-The and My/Mine for the following type of logical expression: e.g., [_DP D [+Def] +N]; as in The-My/Mine book, etc. (This amounts to saying that there is no clear-cut distinction or reference of Possession for the child at this initial stage-1.) There is some preliminary evidence to suggest that the child's early (overgenerated) use of e.g., Mine's (found in my corpus and elsewhere in the literature) may indeed be accounted for in such a manner (cf. Radford 1990: 108ff) [_DP Mine [D 's] N book]].[vii]

(Theory)          Let's pick-up on Hoekstra et al's observation of a D-I correlation (cf. §1.1). Although Hoekstra et al. take the Definiteness Feature of the Subject as the specific feature deficit responsible for the unspecification of DP (Number being utilized as the main deficit of D leading to null subjects/underspecification of IP), we can naturally expand this notion of Definiteness to the +Agreement feature in Pronominal Possessive D(P)s. This extension is made feasible by Abney's (1987) seminal work which argues that possessive nominals are in fact DPs Headed by a null determiner which carries the formal Uninterpretable/+Agr(eement) property. Thus, following Radford (class lectures 1997), (and keeping with the spirit of Abney (ibid.)) a specified DP phrase in (17) below could have one of the following two structures (17a/b) (17c illustrating either the non-functional VP-stage ( i.e., [øAgr] ) or the underspecified IP-stage (i.e., [-Agr] ):

(17)

(17')

We henceforth follow Chomsky (1995: 261) and favor the structure in (17b) over (17a)--where the possessive 's positions within the Head of D where it checks its Agreement properties. In fact, the checking of possessive 's here is an anomaly of sorts. For instance, Radford (p.c.) has come up with the interesting idea that 's must somehow check its personless features with the nominal in Spec. Since Nominals don't typically carry case (cf. the non-contrast of John_a, John_b and John_c--potentially deriving--Nominative_a, Accusative_b and Genitive_c respectively) it's not at all straightforward what kind of case we are considering here. Well, suppose that one way of accounting for the ungrammaticality of e.g., *My's book/Him's book/Her's book etc. (all of which show 's to be in Head of D) is to say that the possessive 's can't attach to a Pronominal with +Agreement/Case. In other words, this amounts to saying that 's checks for Nominal (default) personless-case features (e.g. John [-Per/-Agr]) and is exclusively associated with non pronominals. In a paradoxical sense, +Agr (cf. 17b) might actually mean the formal checking-off of an -Agr personless feature.

We draw our attention here to the utterance John book (=Gen) (cf. 17c) regarding the prefunctional stage-1 below. In (17c), the Head of DP is vacuous. The above token example is taken from my stage-1 and typify this structure. The claim made here is that while early (nonspecified) DPs contain a Specifier and a Complement, their Heads can be void of any morphological material. The reasoning behind the claim that a DP projects here--as opposed to the more traditional NP-analysis for N+N (=Gen) constructions at this stage--is two fold in nature. (i) Firstly, (cf. §1.1) we wish to suggest a two tier class of DPs: a DP>VP (which has a lexical categorial status), and a DP>IP (which has a functional status). (ii) Secondly, the DP-analysis here jibes with the earlier observation that DPs may be initially miscategorized as having lexical category status (cf. Radford: 1990). Recall we were able to account for the overgeneralization of e.g., Mine's book etc. in this way (see note. 20). Furthermore, we may wish to take a hard stance (following Abney) and suggest that once the determiner The is acquired (cf. §1.2.1), a DP must project. Consider the following token examples taken from the prefunctional VP-stage found in my data (files 2-7):

(18)

In the above structure, the Agr feature is [øAgr] absent altogether. However, since +Def projects (the feature being closely associated with Determinacy), the overall DP projects. Such default DPs at this VP-stage match--though for different reasons--their counterpart DPs within adult small clauses--where the Verb in the matrix clause formally assigns Objective Case (via ECM):

(18')

The above amounts to saying that although the child, at the VP-stage, matches the adult skeletal structure of DPs, she fails to realize any of the formal (-Interpretable/Case) feature specifications of the Head (see note 20). This claim suggests that DPs (i.e., Determinacy) may emerge on the scene at the very earliest stage of language development--triggered by the +Definiteness feature. Hence, categorial features along with +Interpretable nominal features (viz., semantically based f-features) are immediately accessible to the child. Case properties [+Agr/-Intrepretable] on the other hand are out of limits for the child at the prefunctional stage--Case can only emerge once the functional IP-stage has been acquired carrying along with it the likes of formal/abstract properties. We are claiming that Subject/Object DPs indeed emerge as an adult skeletal structure at the VP-stage but that there are no specific formal properties attached to the Head D that have to do with checking per se. None of the Determiners in (18) contain their appropriate Case--all of their claims on (Objective) Case are similarly derived via Default. To summarize, the cited Possessive Nominals (DP>VP) are specified as follows:

(19)

We can recapitulate the above distribution of Objective case by addressing the intrinsic asymmetry found between Subjects and Objects. The notion that only the Subject is effected by Hoekstra et al's D-I correlation (cf. §1.1) adds further support to this asymmetry found between Subjects and Objects--namely, the clausal positioning of these two DPs seems to be of some relevance. This has prompted me to reconsider the idea that apparent Accusative DPs (in Object position) do not necessarily abide by the same constraints as their Subject/Nominative DP counterparts. For instance, Chomsky has suggested that the case of a given Expletive (e.g., there) would depend on its counterpart associate DP within the given clause--E.g., expletive there in the sentences:

(20)

would take on its appropriate Case via its associated DP within the respective clause:

(20')

(There is clear evidence in the data that (i) such asymmetries exist and (ii) that such asymmetries arise in nature from syntactic deficits. Recall that in §1.2.2.1, cf. Table 1.3, it was discussed that Accusative Subjects (via default) were acquired much earlier than their Nominative counterpart.) The above observations could be expanded to say that a DP in Object position is of a very different category (regarding aspects of features) than say a DP in Subject position, even though the lexical entry is apparently the same. I take this difference to be of a lexical vs. formal kind respectively--the difference having nothing to do with the lexical item (per se) but rather having everything to do with the item's feature specification. Specifically speaking, a DP--(i) either at a given VP-stage or (ii) infixed within a VP-projection of an otherwise underspecified IP-stage--could theoretically exist without its formal Agreement features being intact, and carry only Objective case (via default). Hence, at the VP-stage, a Subject D(P) may theoretically hold exactly those same feature specifications as its Object D(P) counterpart--viz., only those +Def/phi-features apply--making it difficult to tell whether the D-I Correlation is even operative. (The notion of an Objective DP in Subject position eventually leads to the complete breakdown of the correlation regarding Subj/Obj-DPs in relation to INFL.)

(Data: Stage-1)           As cited above, my data suggests there to be a prefunctional stage where only N+N (=Genitive) constructions and Caseless (clause fragment) My/Mine DPs appear. This first stage, more generally, preexists any form of (Under)specification regarding an IP (i.e., Tense and/or Agreement): there is in all actuality No IP-projection. One means of describing this stage is to call on the notion of The Lexical Deficit Analysis (=LDA) (cf. Schütze: 1997). The LDA basically states that at an early stage of language development, the child may entirely miss out on e.g., Case markings, Inflections, Agreements simply due to the fact that the child has yet to acquire the specific lexical entry (as in morphologically rich languages) or feature (as in English) associated with Case. An altered tactical approach to LDA would be to assume that although the child has indeed acquired the lexical entries for e.g., Possessive DPs e.g., My/Mine, such entries lack properties of case specification and, thus, are to be considered as completely different (lexical) entries altogether--as opposed to their +Agr case marked counterparts My/Mine: (similar to the DP>IP/VP distinctions of The drawn-on earlier in this chapter). (Though, of course, the two sets are homophonic, we shall take the latter [+Agr] to be case marked only when we have sufficient evidence--taken from other means--that Case has been acquired. The essential criterion for determining (Gen) case here is the acquisition of possessive 's coupled with other factors regarding the acquisition of e.g., Nom/Acc Case, etc.) This amounts to the important observation that lexical entries are defined by their bundle-of-features (some features being acquired later than others). The DPs cited below (Table 1.4) mark Head features as [-Agr/+Def] and never fluctuate between [+/-Agr]. Evidence in support of the [-Agr] deficit comes from the observation that Agreements associated with Nominative Subjects, Possessive 's, and Finite Verb constructions +s--forming the "benchmark" criteria as mentioned above--never manifest at Stage-1 (§1.2.2.1-Table 1.2 for statistics on Nom/Acc Case). Compare and Contrast the Tables (cf. 1.4, 1.5 and 1.6) illustrating a bi-model pattern of Inflectional Acquisition:

(21)

Table 1.4 Stage-1 (files 2-8: 1;10-2;4)

(21')

(21")          Token Examples (cf. Table 1.5 above)

The Data drawn from these two tables (above) seem to suggest that Inflectional (Agreement) is gradually acquired until the mastery threshold establishes itself at around 3:0 years: (cf. R. Brown's 90% criteria). The distinction made here between early "Analogous-The-My/Mine constructs (cf. Table 1.4) and those herein above (Table 1.5) is that the early types never incorporate Verbal elements and are said to be analogous to the early spotted Determiner The (among the very first files)--both DPs The and My/Mine here mark for [+Def] only and are said to be unspecified for Agr. It is clear that at the later stage, where IPs are formed which incorporate Agreements, that My/Mine examples now do have access to proper Genitive Case.

Regarding the early usage of The and "Analogous-The"-My/Mine, the two predominate schools-of-thought would characterize their typical structures based on the above examples as follows:

(22)

In (22i), the DP correlates both its Internal D-Spec feature [+Agr] (My= +Gen) and External I-Head feature as [+Agr] (cat= +Nom), whilst seemingly in the absence of any other overt specification for INFL. Tense is unspecified and Agreement may or may not be specified, depending on the account you accept here. (Whatever the case, no overt factors present themselves in such a way as to force a +Agr specification for (22i) outside of theory internal reasons.) We take it that the most parsimonious structure is clearly that of (22ii) where there is no overt need to project +Agr since the DP might just as well be considered as Objective [-Agr/+Def] (carrying the default case). In this sense, the DP is not +Gen though the two structures phonetically resemble each other (see note 20).

Our claim here is that stage-1--where there is no evidence otherwise for INFL--we follow in the spirit of Radford (ms1997) and suggest that overt DPs (in subject/object position alike) may be forced into taking default Acc case since there are no functional categories that could possibly host the moved element driven by checking. In the sense of (22ii), there is no established checking relation within DP just as there is no checking relation within IP: since Agr/T are inert, they have no need to enter into a checking relation. (In the words of Radford (1990: 103), My/Mine are imposters.) We take the Analogy (cf. 21c) that the Determiner The can equate to the pronominal possessive My as a natural extension, given the fact that they both seem to encode purely semantic/+Interpretable features at this stage-1. In the sense of (22ii), The/My pertain to +Interpretable (phi) features having to do with informal semantics of e.g., definiteness/determinacy, and have nothing whatsoever to do with more formal/abstract features such as Case. Hence, the distribution and inter-relation between overt (Pro)nominals can be summarized as follows (Radford: class lectures '97):

(23)

It is clear from this distribution that the early The-My/Mine examples above can't be considered as having true Genitive case since they are considered here as carrying the Default Objective case (cf. 23c).

(Interim Summary of Stage-1)                        One interesting observation that we can deduce from the data thus far is that a correlation, (pointed out to me by Andrew Radford), seems to hold regarding the general acquisition of a wider range of Inflection types. Radford (Radford & Galasso: ms.1998) makes the observation that a (previously unreported) symmetry holds between (i) the development of Subject-Verb structures on the one hand and (ii) Possessive Nominal structures on the other. That is, the data seem to describe an initial (Stage-1) grammar purely based on a Non-Inflectional-Prefunctional Paradigm. The nature of the correlation, (a correlate which clearly speaks to the protracted nature of Inflectional acquisition), could be claimed as being governed-programmed by maturational factors--i.e. functional categories (in this case INFL) are acquired (somewhat) simultaneously given that once the brain can perceive and generalize such formal categories, the "pandora's box" of functionalism opens thus letting-in (potentially all-at-once) all previously blocked formal aspects of grammar. In this sense, Radford's (1990) original thesis that claims for a functional correlation--i.e., that functional categories DP, IP/CP (as a whole) embody a qualitative kind and thus should be triggered to come "on-line" simultaneously, (in a given child), once those factors of maturation are in place--seems to ring true. For clarity, let's recap the general claims being made here:

(24) Table 1.6

Development of Inflection

Occurrence in Obligatory Contexts

                        Age                  3sg-pres +S^*                Poss 'S

                        2;3-3;1             0/69 (0%)                    0/118 (0%)      (=> VP: Stage-1)

                        3;2-3;6             72/168 (43%)              14/60 (23%)    (=> IP/Optional: Stage-2)

(* 0 of 69 (3sg-pers +S) indicates "true" copula/verbal counts cf. Table 1.9--excluding the early 9 counts which were assumed to be of a stereo-typic nature. Poss. +S cf. Tables 1.4 & 1.5)

I believe an underlying and central idea behind Radford's claim can be proposed: specifically being, that a more general composition of Agreement (acting in either Nominal or Verbal domains) seems to be acquired in unitary fashion for the child. In other words, the child's realization of [+Agr] for the two utterance types: e.g. (i) Pat's cough and (ii) Pat's coughing might be identically encoded by the child as +Agr--independent of whether the Agreement mechanism itself is functioning nominally or verbally (respectively). Radford utilizes Kayne (1994: 105) to suggest that the formal aspects of the two Infections (cited above) are one-of-a-kind and both pertaining to IP:

(25)

 

>In this sense, Kayne reduces the generalities behind the two cases of Inflection to a common ancestral Agreement--viz., both inflections ultimately governed by IP (25c).

(Data: Stage-2)           Insofar that we acknowledge an OI-Stage (Stage-2), we put aside theoretical issues and focus rather on empirical content. This section examines if potential correlations emerge between specified vs. undespecified verbal IPs and relevant agreement vs. nonagreements (respectively) regarding Subject/Object D(P)s. The interest lies in seeing if this Developmental Symmetry (as reported above) holds any further correlates for the OI-Stage. Intuitively, one would expect that Spec-features, say within an Object-DP, are independent of Spec-features within IP regardless of underspecification. Namely, the assumed D-I correlation (Hoekstra et al.) is expected to hold configuration only between the Head-features of a (Subject) D(P) and that of its Specifier-features of INFL. However, as witnessed in Kayne's treatment of an "overriding" IP-based Inflection for Verbs and Nouns alike, and coupled with Radford's observation above linking the two relevant inflections to a single "onset-time" in acquisition, matters surrounding configurational dependencies may not be so entirely straightforward. For instance, overall residual effects of underspecification--given that Agr-O and Agr-S are, in theory, essentially composed of the same formal material [+Agr]--might conceivably be spotted "up-and-down" an underspecified tree structure. (Chomsky (1995: 174) asserts that formal phi-features of Agr equally pertain to Agr-O/Agr-S since distinctions between Subject-S/Object-O labeling here are considered as mnemonic devices.) The purely hypothetical notion I am playing-on runs as follows: iff Agreement suffers a general deficit (i.e., [-Agr]) in one phrase of a sentence, say Object within VP, then the Subject within IP, via extension, likewise instantiates some deficit.

The examples below are merely geared to dispel any scenario for an inter-phrasal IP correlate: showing Kayne's Agr-based accounts of Verb/Nominal S-Inflection don't invalidate traditional hypotheses for independency of Spec vs. Head feature specification.

(Poss. Nominals with Copula)            The first clear bit of evidence against any such wide-ranging Agr-based correlate comes from Possessives with Copula Verbs. This stage represents the onset of functional categories where IPs are seen to project at will. Table 1.7 below illustrates that the specifications of lower Object DPs are independent of the upper INFL: i.e., both (higher) Subject/Gen D(P)s and (lower) Object/Gen D(P)s can go either specified or underspecified (independently) within an overall INFL-projection.

(26)            Table 1.7

            Poss. Noms with Copula Verb

                                                                                                Token Examples

                        IP constructions                        a'. It is my t.v. It is your "e". That's your car.

            a. Correct Case:           n.=190             It's Nicolas'. It is a man's paper. What's his-

                                                                        b'. Him eye is broken. Him bike is broken.

            b. Incorrect Case:         n.=40               Where's daddy bike? It is a boy bike. It's-

                                                                        dad turn. Daddy t.v. is broken.What's him-

(26')

The ratio of incorrect-to-correct Possessive Case constructions within an IP Copula phrase is approx. 1:5 (respectively) with an overall 17% incorrect usage for required contexts. The percentage may not seem large at first sight, but keep in mind that our only task here is to dismiss any notion behind a wide-ranging Agr correlation--nothing more. In fact, the 17% usage of incorrect Poss. Noms. within IPs should by no means be cast as insignificant.