Thomas McFadden

Papers for downloading

Many of the papers that are now (or used to be) linked here are part of my dissertation, which can be downloaded from this page. For the most part, the versions in the dissertation are more recent than the ones here, but the versions here are intended to stand alone, whereas the dissertation chapters are not. Please e-mail me if you have questions about the most recent version of any specific paper.

`The distribution of subjects in non-finite clauses: an account wihout Case.' .pdf .ps [see dissertation chapter 8]
Write-up of talk given at PLC 28, February 29th, 2004, submited to conference proceedings.

In this paper I consider the area where syntactic Case does the most work in current syntactic theories, the distribution of subjects in non-finite clauses. The common assumption is that overt DPs are by default not licensed in such positions and thus can only appear when some special mechanism like ECM intervenes. I argue that the facts both within English and cross-linguistically do not bear this out, rather that overt subjects are generally allowed in non-finite clauses. Indeed, the environments where overt subjects are not allowed form a natural class. On this basis I argue that the basic idea of syntactic Case as a licensing condition on DPs is ill-suited to account for the overt distribution of DPs, and that instead we need something more like the EPP, whereby DP movement is driven by the needs of functional heads. This paper consitutes the continuation of the research presented as `Adventures in resolving redundancy' at PLC 26, also available below. A more recent version, incorporating the material from both papers and some additional material, is available as Chapter 8 of my dissertation.

`On morphological case and word-order freedom.' .pdf [see dissertation chapter 5]
Write-up of talk delivered at BLS 29, Berkeley, February 14th, 2003. To appear in conference precedings.
`On the synchronic and diachronic role of case-marking in word order'.pdf
Handout from talk delivered at the Conference on Comparative Diachronic Syntax, Leiden, August 29th, 2003.

In these talks I consider the well-known correlation between rich case-marking and relative word-order freedom. I argue that any attempt to encode a correlation of this type in the principles of the synchronic grammar runs into serious theoretical problems, and furthermore that the empirical data on the correlation do not support an actual implication (whether bi-conditional or one-way), but rather a strong tendency. I conclude that a more traditional explanation based on language use and acquisition can better account for the facts of this phenomenon while avoiding the theoretical difficulties.

`How much of the Germanic strong adjective inflection is pronominal?' .ps .pdf
To appear in Münchner Studien zur Sprachwissenschaft

I re-examine here the old problem of the genesis of the strong adjective inflection in Germanic. The traditional theory is that pronominal endings spread analogically to the adjective paradigm, replacing the older nominal endings, in a process that continued over a considerable period of time, never went to completion, and went further in some dialects than others. I argue that this is an unsatisfying conclusion because it gives no explanation for why the analogy should have happened in some instances but not in others. Instead, I adopt a very strong hypothesis, according to which pronominal inflection was adopted wholesale by the strong adjectives on the basis of the so-called pronominal-adjectives, which had pronominal inflection at a sufficiently early stage. I then argue that nearly all of the adjective endings which are not clearly pronominal can be shown to be ambiguous between nominal and prominal, then argue that the only two clearly nominal forms (N s. f. of the u- and i-stems in Gothic) would have been prime candidates for analogical replacement, and thus their nominal endings may not be original.

`The underlying structures of German inherent Case.' .ps .pdf [see dissertation chapter 4]
Write-up of talk delivered at the joint meeting of the FGLS and the SGL, London, January 3rd, 2003.

This paper argues that objects which are assigned inherent dative case in German are in fact structurally distinct from accusative direct objects, and that it is the structural difference that is responsible for the case-marking, subject-inelligibility and several other syntactic and semantic characteristics according to which they are more like indirect objects than direct objects. Particular analyses are defended for three separate classes of inherent dative verbs which draw explicit parallels to the two classes of double object verbs.

`Adventures in resolving redundancy: Case vs. the EPP' .ps, .pdf, April 30th, 2002. [see dissertation chapter 8]
Write-up of talk delivered at PLC 26, March 3rd, 2002, submitted to proceedings volume.

This paper examines the well-known redundancy in the coverage of syntactic Case and the EPP in the positioning of DPs. A number of recent works have attempted to reduce this redundancy by eliminating the EPP from the theory. I argue that the main assumption on which this work seems to be based, that Case is indispensable because of its role in the morphology, is unmotivated, because in fact syntactic Case and morphological case are quite independent. Thus any attempt to reduce the redundancy between Case and the EPP cannot begin by assuming that Case is untouchable. Indeed, I argue that, if anything, it is syntactic Case that should be eliminated from the theory. I then go on to discuss what this would involve, making some initial (and admittedly speculative) suggestions on how the epirical difficulties that are raised can be dealt with. See `The distribution of subjects in non-finite clauses: an account wihout Case' above (PLC 28) for the continuation of the research presented here.

`The Morphosyntax of Finno-Ugric Case-marking.' Talk delivered at NELS 32, October 20th, 2001. [see dissertation chapter 7]

This is an update of the analysis given for the ordering of Case affixes with plural and possessive markers on nouns in Mordvin and Mari which is proposed in `The Morphosyntax of case and adpositions...' For the talk I concentrated on showing why normal syntactic raising cannot account for the facts, and an analysis with lowering or the like is necessary. There are links here to the handout, .ps, .pdf and to the written version of the paper that will appear in the proceedings (sorry, this one had to be generated in MSWord, so it doesn't look very good). .ps, .pdf.

`The rise of the to-dative in Middle English' .ps [696 Kb], .pdf [70 Kb], November 27th, 2000.
Appeared in D. W. Lightfoot, ed. 2002. Syntactic effects of morphological change. Oxford University Press.

This is a study of the Middle English origins of the ditransitive construction where what is traditionally called the indirect object is marked by the preposition to. Using the Penn-Helsinki Parsed Corpus of Middle English, 2nd edition (Kroch and Taylor 2000), I examine the first appearance of this construction in the texts, its frequency relative to the unmarked double object and the possible orderings of the objects in both constructions. Based on the findings, I propose that the new construction is merely a new morphological realization of the syntactic structure underlying Old and Early Middle English unmarked double objects with the order DO-IO. (Please pardon the appearance. This one had to be done in Word and didn't come through perfectly.)

`The short-stem forms of `go' in Germanic' .ps [320 Kb], .pdf [544 Kb], January 17th, 2001.

The old Germanic dialects show an array of similar but distinct forms with the meaning `go'. Attempting to relate them formally to one another and to forms found in other Indo-European languages has proven extremely difficult and has prompted an incredible volume of work over the last 140 years. In this paper, I examine the facts relevant to the shorter forms, Old High German ga:n/ge:n, Old English ga:n, Old Saxon ga:n Old Frisian ga:n Old Swedish ga: Crimean Gothic geen, which collectively point to a Proto-Germanic doublet *ge:-/gai-, and the theories that have been proposed to clear up the relationship between the two pre-forms and their origins in Proto-Indo-European. In effect, this is a critical review of the considerable literature on the subject, which attempts to integrate the insights gained and determine which of the theories that have been proposed may, in fact, be correct, and which cannot. The longer forms, Gothic gaggan, OHG gangan etc., and their relation to the shorter forms are not treated in this paper.

`Grammaticalization in Choctaw Negation' .ps [68 Kb] , .pdf [33 Kb], May 4th, 2000.

The case system of Choctaw pronominal agreement elements is of the type described variously as active, active-stative or split-intransitive. In such a system (which is contrasted with nominative-accusative and ergative-absolutive) the subjects of intransitive verbs are marked in some cases like the subjects of transitive verbs and in others like the objects of transitive verbs. In Choctaw, the distinction is made primarily according to whether the subject controls the scenario denoted by the verb or not (thus the subject of `run' is marked like a transitive subject, that of `fall' like a transitive patient). Choctaw also has two distinct means for expressing sentence-level negation, one via a specifically negative inflection, the other via the adverb kiyoh. I argue in this paper, based on the form of the adverb, that it is in fact a negative form of the verb ia - `to go', which has been grammaticalized as a periphrastic marker of negation. Based on the semantic distinction reported between the two types of negation and its relation to the encoding of control in agreement markers, I then propose a specific chronological account of how the current alternation could have arisen.

Back to Tom's homepage