4. Final Remarks and Conclusion
(Word Order)       One interesting aspect of Word Order is that it becomes fixed once DASs project. However, in light of this, it seems that SASs and DASs do occur alongside one another (with only the SAS counterpart allowing variable orderings). Thus, it becomes rather difficult to refer to the two modes here as individual stages. In this sense, the eventual correct setting of word order, as partially brought about by the Comp Parameter, is also dependent upon the non-parameterized workings of LCA coupled with DAS (Kayne op.cit.). This is made apparent by the fact that if word order were based purely on Parameterization (for example, via some functional category), then word order should be correctly set for the entire range of XP-structures found a the presumed parameterized stage--given the emergence of DASs (be it for SASs or DASs alike). As we see from the data here, accounts of word order based purely on parameterization don't suffice--this seems to support Kayne's claim for a "Non-Parameterized" LCA.

One additional piece of support in favor of a dual SVO vs. OVS base-generated pattern comes from Sam Epstein (Chomsky 1995: 391 fn108) who notes that Kayne's LCA allows free temporal ordering (either left-to-right, et vice versa). E.g., a satisfied LCA could have any arrangement of (sister) Head-Comp, Comp-Head relations (read-books, books-read freely). It is in this sense that I couple LCA with a Comp-Parameter.

(Underspecification)      There has been some talk that the two prevailing schools-of- thought (as cited within) may eventually be forced together--on principled grounds--by Chomsky's Merger Theory. What I am on about here is the idea that certain features may Merge while others get postponed--irrespective of the category which binds them. In other words, a new brand of classification may be in the making--i.e., categorizing +Interpretable vs. -Interpretable features constructs. For reasons of time and space, I won't go on about in here. But, in what I believe is a sample of things to come, reflect on the idea of traditional Functional Categories (DP, CP) as being able to project at the earliest MLU stage (say, at our stage-1 here) via +Interp(retable) features only (i.e., merge can operate on partial feature sets.) What would one claim about this stage? Since category labels no longer enter into the equation-and we now forfeit the very criteria that once spun old debates on Structural Continuity--the notion of a VP-stage becomes irrelevant. In other words, the issue here is that functional categories (DP, IP, CP) encode both [+Interp] and [-Interp] features (Agr = [-Interp] for IP). In the event that all [-Interp] formal features are lacking at stage-1, would a proto functional category actually be functional or lexical? (See also §5 below).

(Wh-Movement)       (For further arguments against an overall Adjunction account, see Galasso: 1999). Suffice it to say here that any adjunction account for such early, seemingly thematic structures would entail some sort of movement analysis. With such analyses, it becomes exceedingly difficult to account for the wide-spread failure of movement operations and functionalism which usually accompany movement: viz., the cited deficits in Inflectional markings of Tense, Agreement (IP), and Aux Inversion (CP) etc. (See section 2 above for a Non-Inflectional Stage-1). In sum, although children may use Wh+Copula+N constructs as early as MLUw 2.3, such constructs are relegated to being either (i) semi-formulaic in nature or (ii) base-generated strings not involving movement operations where a simple VP projects in order to accommodate the copula verb. An interesting side-note here is that Roeper et al. similarly claim (albeit in reverse effect) that once children come to move Wh-expressions into Spec-CP, they cease to produce subjectless Wh-questions (cited from Radford 1996: 61). What I am suggesting is in fact an opposite 'cause & effect' relation--namely, it is due to the subject surfacing which eventually forces the Wh-element to position into Spec-CP (adhering to Cinque).

(Formulaic Wh)       The three types of early Wh-constructs which all come out as semi-formulaic are the following: (i) What that, (ii) What's/is that, (iii) Where's +N--and are all taken from files 1-8). Somewhat similar Wh+Copula+N constructs which appear starting from file 9 (onward) conversely get considered as maintaining a CP structure via structural continuity (IP truncations notwithstanding). (See Hickey 1993 for criteria).

(A Recap)      The data (above) suggest there to be an initial No-Inflection Stage-1 during which Subject-Verb and Possessor Agr are not marked--a stage characterized by the use of default Objective Possessors/Subjects and the omission of Poss. 'S and third person S. Compare and contrast Stage-1 vs. Stage-2 respectively: (Me car vs. My car; Daddy car vs. Daddy's car; You car vs. Your car; Him car vs. His car; Me/Him go vs. I/He go(es); "What's that" vs. What ø daddy doing/What is daddy doing?) This Optional Stage-2 last until the end of the recordings (file: 25, 3;6), though by file 25, INFL is generally established. The incremental emergence of possessors like His (emerging late in file 24, Table 2.3) demonstrates that different lexical items are acquired at different rates and ages, and likewise, their independent incremental onsets of INFL-related development seem to suggest that the feature specifications involved are lexical-specific (pace Schütze et Wexler), and not dependent upon 'higher-order' phrase structures: IP can be well established for some lexical items e.g., Poss. My = [+Agr] while being totally nonexistent for other items, e.g., Him = [-Agr] for that same period.

(Conclusion)       In conclusion, what we have suggested here is a (manageable) merging of two extreme positions. Firstly, by acknowledging the strongest case scenario as the Null Hypothesis, (Minimise Degrees of Freedom Principle (cf., Hyams 1994)), we appease Continuity Theory. Secondly, and to the delight of the Structure-Builders among us, we slightly 'weaken' the null hypothesis by one degree only: stipulating that in order for any full competence theory to be set in practice, VP-Eligibility must first be exhausted. In other words, if a (prosaic) VP can theoretically handle the relevant material found among the very early stage of language development, then, by virtue of principles of economy of projection (Roeper et al. 1994), it must. However, principles of economy only work in conjunction with this deduced principle of VP-Eligibility--as a natural consequence of Economy--viz., once projected material can no longer be secured within a Q-VP, it must manifest within its destined full-continuity structure. Justification for this approach is self-evident. Following M. Atkinson (op.cit: 32), we can suppose that: [i]f two periods of the data are sufficiently far apart, these sets of data will exhibit rather clear qualitative differences allowing us to construct grammars G₁ (=VP stage) and G₂ (=IP/CP-Optional stage) (respectively) (emphases belong to MA). In this sense, we have shown in general that functional categories do seem to follow at prior stage where (primitive) VPs, along with other lexical categories, function in isolation (i.e., the lexical VP-stage). This was the case with INFL (cf. 2), where a Non-Inflectional Stage-1 preceded a Wexlerian OI-Stage. Likewise, a Wh-VP stage (cf. 3), theoretically, was posited to be in advance of a potential CP-stage. I believe in maintaining these above stances, we come full circle in highlighting traditional differences between the Full Competence vs. Structure-Building schools. This view could be summed-up as follows: from the beginning, there was never much real distance between the two schools (both schools acknowledge that Principles of UG must constrain all possible grammars.)¹⁰ Recent notions of Underspecification have certainly exemplified this--seemingly putting the Structure-Builders on the defensive. However, in full view of the data presented here, the successful 'weakening' of such strong stances has shown once again the (proper) gap between the two schools. By taking the strongest hypothesis (concerning Continuity) as the Null Hypothesis, and then by essentially working backwards and weakening it on a 'need-to-need' basis in order to accommodate the empirical data, we have essentially returned (once again) to the ghosts of bygone orthodoxies: so let's ask the same old familiar question once more (overleaf)--

<<< Do children start-off their grammatical lives with a complete system of syntactic representation...or not? And if so, how can we tell >>>
(emphases belong to Atkinson 1995: 29).

 

5. Residual Problems, Alternative Solutions and Future Research
The following is a highly condensed version of a discussion on Merger Theory found in Galasso (1999). It is presented here merely as a caveat to the aforementioned outstanding issue regarding child formulaic constructs.

In spite of an embarrassing wealth of literature about well know facts on formulaicy in early child speech, (see Brown 1968, Hickey 1993), the notion behind formulaic analyses seems nonetheless to grate on the minds of some researchers today whom esteem to grant the benefit of doubt to the child in any event (and at whatever cost). I myself, among many, personally believe such a granting to be overly optimistic and ill-managed, typically in light of the sort of data the child produces/comprehends at the given stage. However, in acknowledging that there is a potential issue here, and in the event that formulaic accounts do become increasingly unattainable (e.g., there is ongoing research to suggest that children might comprehend Wh-Questions even before they can talk, cf. Radford: pc), I propose below a broad alternative account to default/formulaic grammars which could continue to uphold Structure-Building models of acquisition.

I. (Reconsidering DP)       The traditional idea of functional categories have recently been complicated by the introduction of notions of feature specifications and projections (as cited in Chomsky: 1995, §4 Merger Theory). In Chapter 4 of my Ph.D. Diss., I tease out issues in how we can come to grip with a seemingly functional DP projection in an otherwise (manifesting) Lexical VP stage. In a nutshell, what I discover is that all such DPs used at this hypothetical VP-stage-1 go consistently under/non-specified for formal features [-Interp(retable)] such as Case. By extended hypothesis, I went on to claim (following Radford pc) that such DPs could specify for [+Interp] features only--namely, [+Def(initess)] which would not require any formal checking. This is tantamount to saying that e.g., The within such DPs take-on Case via default and show no other signs of functionalism. I labelled the DPs as either (i) DP>VP or (ii) DP>IP--seeking to express the contrast between DPs which pertain more to the lexical-substantive category VP, contra the traditional classification of a formal functional DP:

      (i)   a. John's car goes =[D/P⁰ [_IP John ['s [car]]] ]^*..goes              [+Agr]    (DP>IP)

            b. John car go =     [D/P⁰ John car]...go                                  [-Agr]      (DP>VP)

(*Bracket structures are simplified to show only the relevant agreement.)

This attempt to redefine traditional notions of categories and phrases in terms of the features that project--viz., [+Interp vs. -Interp]--drastically changes the landscape in how we can assess functional vs. lexical categorial distinction.

II. (Reconsidering CP)         Following in the wake of such underspecified DPs (above), it is not too inconceivable to then propose a similar analysis to underspecified CPs (likewise found at the otherwise patterned VP stage). The parallel [+Interp] feature for CP (as it has to do with Wh-elements) would be Chomsky's notion of a Q-feature (=Interrogativeness). E.g., Chomsky (1995: 289) suggests--that Q is plainly [+Interp]; therefore, like the phi-features of a nominal, it need not be checked (unless strong). Such a treatment of Wh-elements puts them on a par as a variant of D (e.g., What book=DP). Thus, the distinctions drawn-on above concerning D(P)s now have relevance here to C(P)s as well.
Suppose, for Child English, we take Q to be initially set (or unparameterized) to a default 'weak setting' (until further sufficient input on Wh-raising sets it otherwise). Then, Q doesn't raise for checking.
Suppose, for Child English, Q only initially projects this [+Interp] feature. The parallels of the two arguments become ever so apparent--i.e., there is nothing in principle which would prevent us from saying that, like DPs (before them), CPs may also receive a form of default setting.

     (ii)  a. What car go =       [C/P⁰ (VP)...[_DP What car]]^*...go...      [-Agr] (CP>VP)

            b. What car goes =    [C/P⁰ (IP)...[_DP What car]]....goes      [+Agr] (CP>IP)

            c. Whose car      =    [C/P⁰ (DP)...[_IP Who 's car]]        [+Agr] (CP>IP)

This amounts to saying that in spite of a seemingly (straightforward and traditional) CP analysis, such Wh-sentences may be reduced to projecting in one of three base-generated manners via a VP. More specifically, the DP in e.g., What car go may select the following available landing sites: (i) a Spec-VP, (ii) an Adjunct VP, or finally (iii) a Spec CP>VP (with a [+Interp] feature only)--the essential point being here that all three constructs would possibly show no movement. In the latter sense (iii) (exemplifying +Interpretable features here), Roeper & Rohrbacker's CP>VP truncated account could be reinterpreted as ultimately being base-generated.

In full spirit of the proposal sketched above, we then are left with only Agr/T (=IP) to consider. Following Chomsky's (1995, Ch.4) remark that only Agr is left to constitute pure non-Interp features--and having shown how the remaining functional categories DP & CP might be reducible to their substantive feature-projections--we take IP to be the functional category par excellence. As a consequence of no INFL-related material, the feasibility of a CP>DP/VP structure projecting from the earliest MLUw stage may not entirely jeopardize an otherwise seemingly straightforward Structure-Building account of language acquisition. (See Galasso 1999, Chapter 4 for DP & Chapter 5 for CP.)

            III. (Future Research: Issues in Specific Language Impairment)

The protracted nature of language acquisition as drawn up here certainly may aid developmenal linguists and language pathologists interested in accounting for specific areas of delayed grammatical development in some cases regarding Specific Language Impaired individuals (SLI). Whether it be stroke victims suffering from Left/Right Hemispheric damage, or children with Focal Lesions, to instances of Aphasia, I believe a better understanding of the natural language process at hand among normal children can serve as a road-map to uncovering exactly where the linguistic anomalies will surface in the speech output--leading perhaps to better methodologies in therapy. To put this on more concrete grounds, take for example recent reports dealing with in vivo PET studies. Although such early studies--which initially set out to locate and map precise areas of language activity in the brain--have come about with not much success (see Peterson, Fox, Posner, Mintun, and Raichle 1989), there nonetheless have been some sound reports cited to suggest that when a part of language is indeed effected, the effected part pertains to the more formal aspect of language--particularly those categories dealing with Inflectional Morphology and/or Case. This seems to go hand-in-hand with the standard assumption that suggests that Neural Plasticity--which speaks to a critical period of language acquisition: (cf. Lenneberg 1967)--may facilitate recovery more in children than in more mature subjects. In short, this has the flavor of saying that it is those more formal categories of language--i.e., categories prone to maturational development such as Inflection, Agreement paradigms, and Case (cited herein)--which seem to be effected most due to their having a qualitatively different mapping system in the neural make-up of the brain.

Although research concentrating on the effects of Left/Right hemispheric damage is still in its infancy, some early signs suggest that significant delays show up in the child's usage of morphological inflections--including plural, possessive, present progressive, and regular past tense (-ed) forms (Fenson et al. 1993). (See Galasso 1999 paper no. 2 in prep. for further discussion). If by examining the type of errors made by normal children we begin to spot similarities to that of SLI children, we may begin to hypothesis exactly where and how language problems will arise.

In closing, the overall analysis presented here may shed some light on what we know about SLI individuals. The fact that Word Order, INFLectional, as well as Movement Operations seem to be the last of a series of systems to mature due to their abstract nature, it therefore should be of no surprise to find that certain similarities hold with respect to 'what gets missed out where' when contrasting the SLI speech of young-adults. It may very well be the case that some SLI individuals simply have a higher threshold with regards to what amount of formal speech input is required in order to trigger a certain formal system into production. This amounts to saying that a five-to-six year old typical SLI child may be in parallel development to that of a two-to-three year old normal child. Perhaps, it could only be such an analysis that ultimately accounts for scattered reports which tend to show that the speech of young SLI children eventually does catch-up to the appropriate normal level of linguistic production.

The main aim of the second paper is to focus primarily on the Quantitative/Qualitative measurements behind an SLI individual--using what has been found and discussed in this paper as a theoretical starting point from which to precede.

Acknowledgments
This paper represents a rather simplistic view of my Ph.D. Diss. (1999). I am grateful first and foremost to Andrew Radford (Essex) (my mentor and Ph.D. supervisor) for his valuable comments on earlier drafts of this paper. I also thank Sam Featherston for his reading through an earlier draft, and Harald Clahsen (Essex) for lending me access to his data-base coding convention. I thank Hiro Matsumoto for our lengthy discussions on Kayne.

<< Back to Index

References

Atkinson, M. (ms.1995) Now, hang on a minute: some reflections on emerging orthodoxies. Essex Research Reports In Linguistics: Vol.7 pp. 29ff. University of Essex.

Bickerton, D. (1990) Language & Species. The University of Chicago Press: Chicago

Bowerman, M. (1973) Early Syntactic Development. Cambridge University Press: Cambridge.

            (1990) 'Mapping Thematic Roles onto Syntactic Functions: are children helped by innate linking rules?' Linguistics 28. pp.1253-1289 Walter de Gruyter.

Brown, R. (1973) A First Language: The Early Stages. George Allen and Unwin: London

Chomsky, N. (1981) Lectures on Government and Binding. Foris, Dordrecht Press.

             (1995) The Minimalist Program. MIT Press: Cambridge, Mass.

Cinque, G. (1990) Types of A'-Dependencies. MIT Press: Cambridge, Mass.

Clahsen, H. (1986) 'Verb Inflections in German Child Language: acquisition of agreement markings and the functions they encode'. Linguistics, 24: 79-121.

Clahsen, H.; Eisenbeiss, S.; Penke, M.(1995) "Lexical Learning in Early Syntactic Development". In H. Clahsen (Ed.) Generative Perspectives on Language Acquistion. Benjamins: Amsterdam

Felix, S.(1984) 'Maturational Aspects of Universal Grammar'. In A. Davis, C. Criper and A. Howatt (eds). Interlanguage. Edinburgh Univ. Press: Edinburgh.Fenson, L. (1993), Dale, P.S., Reznick, J.S., Thal, D., Bates, E., Hartung, J. Pethick, S. and Railly, J. The MacArthur Communicative development Inventories. San Diego.

Galasso, J. (1994-1998) The N-Corpus. (Personal Data-Base compiled from 8,000+ analyzable sentences. Coding convention was established by H. Clahsen (1984)       University of Essex).

Galasso, J. (1999a) The Acquisition of Functional Categories: A Case Study. Unpublished Ph.D. Dissertation. University of Essex.

           (1999b) A Working Paper on Second Language Acquisition Research: Some   Notes on Theory and Method. Unpublished Manuscript. San Diego State University Summer, 1999 (Class Lectures taken from Linguistics 550).

            (In prep) Grammatical Development and Language Disorders (Paper no. 2). Unpublished manuscript. Long Beach City College.

Hickey, T.(1993) 'Identifying Formulas in First Language Acquisition'. J. of Child Language 20: 57-125.

Hoekstra, T., Hyams, N., Becker, M. (1996a) The role of the specifier and finiteness in early grammar. Paper presented to the child language seminar: Reading, April 1996.

           (1996b) The underspecification of number and the licensing of root infinitives. Paper presented to the Boston University Conference on Language Development.

Hyams, N. (1995) 'The Underspecification of Functional Categories in Early Grammar'. In Harald Clahsen (ed) Generative Perspectives on Language Acquisition. Benjamins: Amsterdam pp. 91-127

Kayne, R. (1994) The Antisymmetry of Syntax. MIT Press: Cambridge, Mass.

Klima, E.S., and Bellugi, U. (1966) 'Syntactic Regularities in the Speech of Children'. Psycholinguistic Papers. (eds) J. Lyons & R. Wales. Edinburgh Univ. Press.

Lenneberg, E. ( 1967) Biological Foundations of Language. Wiley Press: New York.

Peters, A. (1983) The Units of Language Acquisition. Cambridge Univ. Press: N.Y.

            (1995) 'Strategies in the Acquisition of Syntax.' In P. Fletcher & B. MacWhinney (eds). The Handbook of Child Language pp.462 ff. Blackwell: Cambridge, Mass.

Peterson, S.E., (1989) Fox, P.T., Posner, M.I., Mintun, M., and Raichle, M.E. 'Positron emission: tomographic studies of the cortical anatomy of single-word processing'. Nature 331, 585-9

Pierce, A. (1989) On the emergence of syntax: a crosslinguistic study. Diss., MIT.

Radford, A. (1990) Syntactic Theory and the Acquisition of English Syntax. Basil Blackwell: Cambridge, Mass.

            (1994a) Syntactic Theory and the Structure of English. Cambridge Univ. Press.

           (1994b) 'The Syntax of Questions in Child English'. Journal of Child Language: 21. pp.211-236. Cambridge University Press.

            (1995) 'Phrase Structure and Functional Categories'. In Paul Fletcher and Brain MacWhinney (eds). The Handbook of Child Language. pp.483-507. Blackwell: Cambridge, Mass.

            (1996) 'Towards a Structure-Building Model of Acquisition'. In H. Clahsen (ed) Generative Perspectives on Language Acquisition. Benjamins: Amsterdam.

Radford, A. and Galasso, J. (1998ms) Children's Possessive Structures: A Case Study. University of Essex.

Rizzi, L. (1986) 'Null Subjects in Italian and the theory of pro.' Linguistic Inquiry 17 pp.501ff

Roeper, T. 'The Role of Merger Theory and Formal Features in Acquisition'. In Clahsen (ed) Generative Perspectives on Langauge Acquisition. Benjamins: Amsterdam.

Roeper, T., and Rohrbacher, B. (1994) Null subjects in early child English and the theory of economy of projection. (ms. to appear in Linguistic Inquiry).

Schütze, C. (1997) INFL in Child and Adult Language: Agreement, Case and Licensing. Diss., MIT.

Tsimpli, I.M. (1992) Functional Categories and Maturation: The Prefunctional Stage of Language Acquistion. Diss., University College London.

Wexler, K. (1994) 'Optional Infinitives, head movement and the economy of derivation'. In D. Lightfoot and N. Hornstein (eds) Verb Movement. Cambridge Univ. Press p305f

Schütze, C. & Wexler, K. (1996) 'Subject Case Licensing and English Root Infinitives'.   In A. Stringfellow, D. Cahana-Amitay, E. Hughes and A. Zukowski (eds) Proceedings       of the 20th Boston University Conference on Language Development: Vol.2. Cascadilla Press: Somerville, Mass. pp.670-681.

Tonoike, S. (1995) 'Japanese as an OVS Language'. In S. Haraguchi & M. Funaki (Eds)    pp. 105-133

Wakefield, J. & Wilcox, M.J. (1995) 'Brain Maturation and Language Acquisition: A       theoretical model and preliminary investigation'. Proceedings of the 19th Annual       Boston University Conference on Language Development. 2 vols,(eds) by D. MacLaughlin & S. McEwen. pp.643-654. Cascadilla Press: Somerville, Mass.

Winjen, F. and Bol, G. (1993) "The Escape from the Optional Infinite Stage'. Papers in Experimental Linguistics: Univ. of Groningen: pp. 239ff.

Section 1 (on Kayne) was presented at the Annual Research Presentation at the University of Essex (20, May 1998).

10 Here, I speak against notions of 'wild grammars' (Felix 1984). Moreover, tactics for defining absolute feature-nonspecificity (for the proposed stage-1) could equally be cast within Structure-Building models--with more traditional questions like 'How do category/phrase types emerge and develop?' being replaced by the now relevant question--'How do particular feature types develop?'