NORMS Workshop on Determination
Tromsø, 18 – 19 March 2009
|
|
Abstracts
| Muhsina Alleesaib (Paris/Berlin): Insights on the 'definite' determiner LA in Mauritian |
Mauritian has in common with other French-derived Creoles a postnominal 'definite' determiner, LA, illustrated in (1), which has evolved from the French locative morpheme là (illustrated in (2)).
| (1) | a. | Panye | la | lour. | (2) | a. | Je | préfère | cette | robe- | là. |
| Basket | LA | heavy | 1SG | prefer | DEM | dress | LA | ||||
| 'The basket is heavy.' | 'I prefer that dress.' | ||||||||||
| b. | Rober | ti | la. | b. | Robert | était | là. | ||||
| R. | PAST | LA | R. | be-IMP | LA | ||||||
| 'Robert was here.' | 'Robert was here.' | ||||||||||
The term 'definite' is used in Syea (1996), Bollée (2000) and Déprez (2006,2007), but not by Robillard (1993), Zribi-Hertz (2002) and Zribi-Hertz and Glaude (2007). These authors highlight spatial dimension and/or anaphora as characteristics of LA in Mauritian and in Haitian. I will argue that while LA occupies the D-head of the DP, its distribution differs from that of D-heads in other languages. Furthermore, while the use of LA in Haitian is relatively well-documented in generative grammar, this is not necessarily the case for Mauritian. I present some differences between the use of LA and that of the definite determiner in French, based on Hawkins (1978). LA is used for subsequent mention of an already introduced DP, or it can be used in associative anaphora, characteristics it shares with French le.
Le but not LA can occur in DPs denoting non-specific referents, whether in generic or non- generic sentences. (3) illustrates the latter case.
| (3) Context: The speaker needs meat. | ||||||||||||
| a. | Je | vais | chez | le | boucher. | b. | Mo | pe | al | kot | bouse | (*la) |
| 1SG | go-PST | PREP | LE | butcher | 1SG | PROG | go | PREP | butcher | LA | ||
| 'I'm going to the butcher.' | ||||||||||||
LE, but not LA, can be used with (contextually) unique referents.
| (4) Context: - What's new ? | |||||||||||
| a. | Le | Premier | Ministre | a | démissionné. | b. | Premye | Minis | (#la) | inn | demisyone. |
| LE | Prime | Minister | AUX | resign | Prime | Minister | LA | PERF | resign | ||
| 'The Prime Minister has resigned.' | |||||||||||
Incidentally, the Mauritian examples (4b) also shows the same 'definite' reading can be obtained in the absence of LA, in contrast to (4a) where le is obligatory, which suggests that a 'definite' reading is not strictly correlated to morphological marking in Mauritian. It has also been suggested that LA corresponds to the demonstrative in French, but Mauritian has its own demonstrative marker (sa ... la) whose uses do not always overlap with those of LA. By examining the properties of DPs with LA, I attempt an analysis as head of a DP should be complemented by an in-depth study of its distributional properties.
Selected references:
Déprez, V., 2007, Probing the structuring role of grammaticalization: nominal constituents in French Lexifier Creoles, JPCL, 22: 2, pp. 263–307
Robillard, D. de, 2000, 'Plurifonctionnalité de(s) 'la' en créole mauricien: catégorisation, transcatégorialité, frontières, processus de grammaticalisation,' in L'Information grammaticale, 85
Hawkins, J.A., 1978, Definiteness and indefiniteness, London: Croom Helm
Syea, A., 1996, 'The development of a marker of definiteness in Mauritian Creole', in P. Baker and A. Syea (eds.) Changing meanings, changing functions: papers relating to grammaticalization in contact languages. London: University of Westminster Press.
Zribi-Hertz, A., Glaude, H., 2007, 'Bare NPs and deficient DPs in Haitian and French: from morphosyntax to referent construal, in M. Baptista & J. Guéron (eds.), Noun Phrases in Creole languages, Amsterdam: Benjamins, pp. 265-298
| Merete Anderssen (Tromsø, CASTL/NORMS): Multiple indefiniteness markers and adjectival inflection in a North Norwegian dialect |
While the occurrence of multiple definite articles in Modern Greek is a well-known and well-studied linguistic phenomenon (cf. e.g. Alexiadou & Wilder 1998, Leu 2007 and Lekakou & Szendrői 2007), the existence of multiple indefinite articles has received considerably less attention in the literature. Indeed, in some accounts of so-called determiner speading in Modern Greek, poly-indefiniteness is seen as an impossibility. This paper aims to consider this question by studying multiple indefiniteness markers in a North Norwegian dialect spoken on the island of Senja.
Polyindefiniteness in the Senja dialect is seen in indefinite noun phrases involving one or more adjectives, as illustrated in (1a) - (1c). The presence of all except the first article is optional, as indicated by parentheses. Note, however, that that recursive indefiniteness is not permitted with non-intersective and nationality adjectives, as shown in (1d). This is similar to what has been observed for determiner spreading in Modern Greek (Alexiadou & Wilder 1998).
| (1) | a. | Ei | stor | (ei) | fin | (ei) | seng | |
| a.fem | big.M/F | (a.fem) | fine.M/F | (a.fem) | bed | |||
| b. | En | stor | (en) | fin | (en) | gutt | ||
| a.mas | big.M/F | (a.mas) | fine.M/F | (a.mas) | boy | |||
| c. | Et | stor-t | (et) | fin-t | (et) | hus | ||
| a.neu | big-neu | (a.neu) | fine-neu | (a.neu) | house | |||
| d. | En | påstått | (*en) | morder. | ||||
| a.mas | alledged | a.mas | murderer |
| (3) | a. | Ei | stor-e | fin-e | seng |
| a.fem | big-e | fine-e | bed | ||
| b. | En | stor-e | fin-e | gutt | |
| a.mas | big-e | fine-e | boy | ||
| c. | *Et | stor-t-e | fin-t-e | hus | |
| a.neu | big-neu-e | fine-neu-e | house | ||
| d. | *Et | stor-e | fin-e | hus | |
| a.neu | big-e | fine-e | house | ||
| e. | *En | påstått-e | morder. | ||
| a.mas | alledged-e | murderer |
Despite the apparent similarity between the -e suffix and the weak adjectival form, there is reason to believe that they are not equivalent, because the two occur with different pitch accents; the weak adjectival form has pitch accent 2 (low-high-low), while the -e suffix that occurs on the adjective in indefinite noun phrases has pitch accent 1 (high-low). Thus, it would appear that the -e suffix is distinct from the weak adjectival form. The recursive article and the -e suffix, on the other hand, are in complimentary distribution, as illustrated in (4), which could be taken as an indication that they are two expressions of the same phenomenon.
| (4) | Ei | stor-e | (*ei) | fin-e | (*ei) | seng |
| a.fem | big-e | (a.fem) | fine-e | (a.fem) | bed |
The present paper discusses the various restrictions on the distribution of polyindefiniteness in the Senja dialect and its interaction with the -e suffix on strong adjectives. It will be argued that the main characteristic of these constructions is that they are highly descriptive, and we will ask whether a reduced relative clause analysis is the best way to capture these structures.
| Kersti Börjars and John Payne (Manchester): Definiteness in the Scandinavian noun-phrase: an LFG account |
In this paper, we present an analysis of definiteness in the Scandinavian noun phrase using the framework of LFG-OT. Within this analysis, the primary differences between Scandinavian languages result not from the lexicon, but from different rankings of constraints.
Following the development of the DP hypothesis, most work in generative syntax assumes that functional categories in the noun phrase are justified by any exponence of the corresponding functional feature. For example, the Danish definite article det, the Danish definite inflection -t and the Welsh definite clitic yr= represent different kinds of exponence of the feature [+DEF]. All these elements are then taken to justify a syntactic category D, which then qua head projects to DP. In LFG, syntactic representations are required to make formal distinctions between structural exponence, where a D node is postulated, and morphological exponence, where it is not. It is the feature [+DEF] itself which must feed the functional/semantic level of representation. A definite noun phrase may in principle have one or more exponents of [+DEF], as long as consistency is observed.
In our analysis, the position, type and frequency of exponence is determined by constraint ranking. In particular, we make use of an alignment constraint that requires any exponent of definiteness to be aligned with the left edge of the noun phrase. A high ranking of the left edge constraint, in conjunction with a constraint barring displacement of the head noun, forces the use of articles when a definite noun is separated from the left edge by an adjective, e.g. Danish æblet (apple.DEF) 'the apple', det store æble (the big apple) 'the big apple', but *store æblet. Weak adjectives like store cannot represent a definite noun phrase in isolation, do not bear the [+DEF] feature and do not satisfy alignment. When no adjective is present, the noun itself is required to carry morphological marking of definiteness, and satisfies alignment.
Scandinavian noun phrases are also well-known to exhibit cases of "definiteness agreement" in which structural exponence of definiteness at the left edge is accompanied by morphological exponence on the head noun. In Swedish det stora äpplet (the big apple.DEF) 'the big apple', the requirement for multiple marking clearly outweighs the requirement that any exponent of definiteness should be aligned with the left edge. The differences between Swedish and Danish emerge straightforwardly from alternative rankings of the multiple-marking constraint.
| Tabea Ihsane (Geneva): Determination revised: languages with and without determiners |
Background and analysis: This paper explores a view of determination which differs radically from the traditional one which attributes a feature [±definite] to a single position (Ihsane and Puskás among others). The claim is that definites and indefinites come in different shapes, i.e. their respective left peripheries do not comprise the same layers. Consequently, indefinites and definites belong to different categories. A predicate thus selects an indefinite or a definite structure, depending on the context (familiarity ...). These structures, we argue, are the same across languages, i.e. in languages with articles (French), languages with articles and bare nouns (English) and determinerless languages (Finnish). The analysis further allows us to distinguish between several types of indefinites and definites. The former, for instance, may represent three categories, depending on there being one, two or three layer(s) in their left periphery.
Data: The category of a dog in (1) depends on its (non)referential status: it is a Property Phrase (PropP) when it is indefinite and means any-N (1)a and a Speaker-Reference Phrase (SRefP, which comprises a PropP layer) when it is indefinite and encodes a dependent reference (in contrast with definites) (1)b. In (1)c, the dog is a Rigid-Reference Phrase (RRefP) as it is definite. What kind of complement occurs in a sentence is a matter of selection.
| (1) | a. | I'm looking for [PropP a dog]. (any dog) |
| b. | I'm looking for [SRefP a dog]. (Rex, my dog) | |
| c. | I'm looking for [RRefP the dog]. |
The articles in (1) are not generated in PropP, SRefP or RRefP but in the inflectional domain. They move to the highest projection to licence the various layers of the nominal structure (cf. Borer 2005). The result for (1)a is (2), where FPcount leads to an individuated reading. The analysis also applies to bare nouns. In (3), letters is a Property Phrase and in the absence of determiner, it is the noun which moves to PropP (3)b. In the same logic, we expect the account to extend to determinerless languages like Finnish (4).
| (2) | PropP | NumP | FPcount | NP | |||
| a dog: | a | <a> | <a> | dog | |||
| (3) | a. | Mary wrote letters. | |||||
| PropP | NumP | NP | |||||
| b. | letters: | letters | <letter-s> | <letter> | |||
| (4) | ...kirjoitti kirjeitä | Kiparsky (1998) | |||||
| ... wrote letters OR wrote the letters | |||||||
In (4), the object may be interpreted as indefinite or definite. In the former case, it is analysed like letters in (3) and in the latter case, it is a RRefP, on a par with the dog in (1)c. In both readings, kirjeitä moves through the structure (maybe as an XP) to the highest layer, PropP or RRefP respectively. The indefinite or definite interpretation comes from the structure selected.
Scandinavian languages: The above account raises many questions in particular for the analysis of Scandinavian languages attesting what is known as 'double-definiteness' (5).
| (5) | den | gule | skjort-a | Norwegian | |
| def.sg | yellow | shirt-def.fem.sg | 'the yellow shirt' | (Julien 2003) |
If the determiner moves through the inflectional structure as proposed, it would be interesting to see if 'double-definiteness' couldn't be a case of 'multiple spell-out': the definite marker on the noun would be a trace of the determiner which is pronounced. This would involve NP movement to the specifier of the lowest functional projection, for reasons to be determined.
| Janne B. Johannessen, Signe Laake and Arne Martinus Lindstad (Oslo, NORMS): Dative in Norwegian – evidence from three dialects |
Dative case is a feature that does not belong in any of the two standard written varieties of Norwegian, bokmål and nynorsk, but which is well known to exist in many dialects. The dative category has only survived on semantically definite categories: preproprial articles, pronouns and definite form of nouns, in contrast to other inflectional categories, and in contrast to Old Norse. Thus, dative is a category that belongs to the determinative system of nominals. Dative case is triggered in the complement position of certain verbs (indirect object of ditransitives and benefactive objects), adjectives and prepositions. A couple of examples are provided in (1):
| (1) | a. | ja ha ru ha ru vøre nå åt sæter-n du nei i høst? | Alvdal: speaker 04 |
| yes have you have you been anything at pasture-Dat.fem.sg you no this autumn | |||
| 'So, have you spent any time at the pasture this autumn?' | |||
| b. | så jekk e tre daga i vek-un då i småskul-un... | Vang: speaker 01 | |
| then went I three days in week-Dat.fem.sg then in small.school-Dat.fem.sing | |||
| '... then I went three days a week to the elementary school...' | |||
We have two objectives with this talk, one empirical and one methodological:
First, we look at three dialects that are, by traditional assumption (see e.g. Faarlund 2000), thought to have dative, and check to which extent this is supported in actual speech. These are the dialects of Alvdal, Skreia and Vang (in Valdres). We show that dative forms persist in the spontaneous speech of at least some speakers from all three dialects. There are, however, considerable differences in the extent to which this grammatical feature is put to actual use, both among speakers of the same dialect, and interdialectally. Dative is maintained as a grammatical category in all three dialects, but in a much stronger degree among the older speakers, showing that the feature is perhaps on the verge of becoming extinct. This is supported by the fact that we have only found dative documented after certain prepositions, and there is variance between the speakers, cf. (2), an excerpt from a conversation between two informants (age above 50):
| (2) | a. | ... ja va de e snakke mæ a Gunnar om dagen | Alvdal: speaker 03 |
| ... yes was it I talked with him.Dat Gunnar on the.day | |||
| '... yes, what was it I talked with Gunnar about the other day...' | |||
| b. | ... ru snakka me n Gunnar om nå så ... | Alvdal: speaker 04 | |
| ... you talked with him Gunnar about something so | |||
| '... you talked with Gunnar about something so...' | |||
(To preserve anonymity, the male name Gunnar has been arbitrarily chosen in these examples.) The preproprial article a in (2a) indicates dative, while n in (2b) is the neutral option. Note that the preposition med 'with' is the same in both examples.
In addition, dative seems to be more prevalent in Vang than in the other two dialects. This might be due to the Vang dialect having preserved a richer inflectional system than the other two in the first place, and it is also grammatically more distinct from the standard language than the other two dialects, thus the pressure from the standard variety might not be felt that strong in this area.
Our second task is to compare two sources of data obtained from the informants. The first source is conversation and interview data, more specifically spontaneous speech recorded, processed and put into a searchable corpus. The second is based on speaker intuitions, i.e. the informants' evaluation of test sentences presented to them in a questionnaire. There is a certain degree of discrepancy between what the informants report in the questionnaire and their actual linguistic practice. One example are young speakers that accept examples like (2a), while they never produce such examples themselves, always resorting to the (2b) variant. This raises important questions about the methodology used in data collection, and shows that the notion 'grammar' (of an individual) in the technical sense of the word might be a vague concept when applying different methodologies. It also accentuates the necessity of having access to different kinds of data when doing fieldwork.
| Tom Leu (Yale): Detecting the adjectival article in determiners, quantifiers, and agreement alternations |
There are a number of environments in which we are - on standard assumptions - sometimes surprised by the appearance of a definite marker morpheme, e.g. German d-. These environments include (i) the pre-adjectival position in definite DPs, which has been discussed as part of "double definiteness" for Scandinavian, as "determiner spreading" for Greek, as "adjectival article" for Slovenian; and (ii) certain "determiners/quantifiers" such as definite demonstratives, e.g. th-is, and German beid- 'both' and jed- 'every', which will be argued to spell out adjectival structures.
Secondly, in German the presence of a pre-adjectival definite marker determines the location of the adjective relative to the strong agreement head (gut-er Wein 'good wine' versus d-er gute Wein 'the good wine'), in a way comparable, as I will suggest, to the alternation in the German tensed clause between a clause initial complementizer dass and V2 (viewed as a sort of vP first).
An attempt to unify these observations suggests the possibility that the adjectival determiner is a left peripheral head in the extended projection of the adjective, xAP: e.g. SWE [ [ det stora ] huset ], GER [ [ der lustige ] Film ]. This in turn suggests a closer assimilation of the higher functional fields of (i) the clause, (ii) the noun phrase, and (iii) the adjectival phrase.
| Erik Schoorlemmer (Leiden): Germanic definiteness marking: morphological variations on the same syntactic theme |
Introduction: Within Germanic languages, there is an intriguing contrast with respect to whether definiteness marking is sensitive to adjectival modification. In German, Dutch and English (henceforth group 1), definiteness is marked by a freestanding definite article irrespective of the presence of an Attributive Adjective (AA): the (old) man. In Swedish (Sw), Norwegian (No), and Faroese (Fa) (henceforth group 2), definiteness is marked differently depending on the presence of an AA. In unmodified DPs, definiteness is marked by a suffix: hus-et 'house-the' (Sw.). However, definiteness is expressed twice in definite DPs with AAs, both by a suffix on the noun and a freestanding article, det stora hus-et (the big house-the) (a.k.a. double definiteness (Delsing 1993, Julien 2005). Despite these two different patterns, I argue that the structure of definite DPs is identical in both language groups. The two different patterns of definiteness marking are attributed to a small morphological difference.
Unmodified definite DPs: I propose that unmodified definite DPs have the structure in (1). The only difference between group 1 and 2 is morphological. In the postsyntactic morphological component (Halle & Marantz 1993, Embick & Noyer 2001), the article that is inserted in D is a free morpheme in group 1 and a suffix in group 2. Because of this, the article in group 2 undergoes Local Dislocation in the sense of Embick & Noyer (2001) to the right of the noun, cf. (1c), while this is not the case for the article in group 1, cf. (1b). Local Dislocation is supported in this case by the observation that some nouns do not take the definite suffix, like den studerende 'the student', cf. Hankamer & Mikkelsen (2005).
Modified definite DPs: I argue that in Narrow Syntax the AA is merged later than the definite D, cf. (2a). This is supported by the sensitivity of the adjectival agreement to definiteness as displayed by the distinction between strong (Sw. ett stor-t hus) and weak adjectival inflection (det stor-a huset). On the null hypothesis that adjectival agreement is licensed by Agree (Chomsky 2000, 2001), this shows that the AA must c-command the definite D and hence must be merged after D. However, this puts the AA outside the semantic domain of the definite D, contrary to its interpretation. In order to put the AA back in the semantic domain of D, D undergoes Internal Merge to a position above the AA, cf. (2b). At the CI-interface, the higher D-copy is interpreted yielding the correct interpretation, cf. (2c).
In the morphological component, the definite suffix of group 2 languages is inserted in both the higher and lower D-copy, cf. (3a). The lower D-copy is suffixed on the noun via Local Dislocation, just like in unmodified definite DPs, cf. (3b). The higher D-copy is however not adjacent to a nominal host. Its nominal suffixal property can therefore not be satisfied by Local Dislocation. At last resort, a dummy host d- is inserted (cf. d-support in Roehrs 2006, Santelmann 1993), cf. (3c). Since the lower D-copy undergoes morphological reanalysis as part of Local Dislocation and morphological reanalysis bleeds Chain Reduction (Nunes 2001), none of the two D-copies can be deleted. This yields the double spell-out of the definite article in group 2 languages.
In group 1 languages, the definite article inserted in the two D-copies is a free morpheme, cf. (4a). Therefore, unlike in group 2 languages, Local Dislocation or another form of morphological reanalysis is not triggered. Hence, Chain Reduction (Nunes 2001) applies; deleting the lower D-copy, see (4b). In this way, the absence of article doubling in the group 2 languages receives a morphological account.
The patterns of definiteness marking in Danish and Icelandic can be accounted for under this analysis by minor variations of operations in the morphological and syntactic components.
| Tania Strahan (Reykjavík, NORMS): Definiteness and Norwegian pronouns |
Despite the quantity of research done into Norwegian personal pronouns, they remain a fascinating object of study. They can occur 1) obligatorially or optionally as PREPROPRIAL ARTICLES before names (Delsing, 1993, Longobardi, 1994); 2) as DETERMINERS (Faarlund et al., 1997), for example the PSYCHOLOGICALLY DISTAL DEMONSTRATIVES of the Oslo dialect (Johannessen, 2006, 2008, etc), contrasting with other, spatially oriented demonstratives; and 3) they can be MODIFIED BY PPS (Hestvik, 1992, Faarlund et al., 1997) and RELATIVE CLAUSES (Faarlund et al., 1997, Vik and Killingbergtrø, 2002); to name but some of the more interesting uses.
Hestvik (1992) uses this last point to argue that pronouns in Norwegian are Ns and not NPs, and can therefore can be modified. However, the approach taken by Faarlund et al (1997) seems even more straightforward. There it is argued that this last use of pronouns, ie those are modified by a relative clause or a prepositional phrase, should (also) be considered determiners, since, among other things, they have no object form (compare de/dem and han/ham in these cases: Jeg så dem/de mennene 'I saw them/those men' and Jeg så ham/han mannen 'I saw him/ that man').
In addition, while pronouns are often considered to have specific reference (Lødrup, 1982, Johannessen, 2008, Strahan, 2008, among others), determiners generally speaking have non-specific reference (Faarlund et al., 1997). This is illustrated in the Norwegian Reference Grammar with [han + relative clause] and [den + relative clause] examples which highlight this contrast. However, it has also been noted that the construction [han + relative clause] may also have non-specific reference, eg Vik and Killingbergtrø's (2002: 1426) entry on han in the Norsk Ordbok Han som ikkje vil høyra, får føla 'He who won't listen, can feel' (cited in Johannessen, 2008: 70, fn).
In Strahan (2008), I suggested a lexical account of the pronoun demonstrative (ie Johannessen's psychologically distal demonstrative and not Delsing's preproprial article). This account attempted to explain two notable facts. Firstly, that the pronoun demonstrative in phrases like han mannen always has a specific referent, in all of the Scandinavian languages that have this construction, (ie Norwegian, Danish, non-Standard Swedish and possibly Icelandic). Secondly, that in all of the Scandinavian languages that have this construction, the accompanying noun must be definite, whether the variety is a double-definiteness variety (Norwegian, non-Standard Swedish) or not (Norwegian, Danish, Icelandic). In this account, all pronouns are lexically specified as [+SPECIFIC]. However, in light of examples like that pointed out by Vik and Killingbergtrø, this account cannot be entirely accurate.
Continuing my original Lexical-Functional Grammar analysis, and following Andrews (2008), I assume that the functional-structure, fed by the lexicon, is only an intermediary level of structure, upon which semantic assembly instructions act. Using Glue language, I propose a rule which results in pronoun demonstratives being interpreted as specific only when they occur with an NP in the syntax. In other cases, the interpretation of specificity is left for a later level of processing.
References
Andrews, Avery. 2008. The role of PRED in LFG + Glue. Paper presented at International Lexical-Functional Grammar Annual Conference 2008, University of Sydney. www.lingfest.arts.usyd.edu.au/LFG/andrews-a_14.pdf.
Delsing, Lars-Olof. 1993. The internal structure of noun phrases in the Scandinavian languages, University of Lund: PhD thesis.
Faarlund, Jan Terje, Lie, Svein, and Vannebo, Kjell Ivar. 1997. Norsk Referansegrammatikk. Oslo: Universitetsforlaget.
Hestvik, Arild. 1992. LF Movement of Pronouns and Antisubject Orientation. Linguistic Inquiry 23:557-594
Johannessen, Janne Bondi. 2006. Just any pronoun anywhere? Pronouns and 'new' demonstratives in Norwegian. In A festschrift for Kjell Johan Sæbø - in partial fulfilment for the requirements for the celebration of his 50th birthday, eds. Torgrim Solstad, Atle Grønn and Dag Haug, 91-106. Oslo: Unipub forlag.
Johannessen, Janne Bondi. 2008. Psykologiske demonstrativer 'Psychological demonstratives'. In Språk i Oslo: Ny forskning omkring talespråk 'Language in Oslo: New research in spoken language', eds. Janne Bondi Johannessen and Kristin Hagen, 63-77. Oslo: Novus Forlag.
Lødrup, Helge. 1982. De og dem i dialekten på Oslo vestkant 'De and dem in the dialect of West-end Oslo'. Norskrift 37:54-58
Longobardi, Giuseppe. 1994. Reference and proper names: A theory of N-movement in syntax and Logical Form. Linguistic Inquiry 25:609-665
Strahan, Tania E. 2008. Sjå på han mannen! On the definiteness and specificity of Scandinavian pronoun demonstratives. In Nordic Journal of Linguistics, eds. Lars-Olof Delsing and Kersti Börjars, 193-226.
Vik, Gudrun Dahler, and Killingbergtrø, Laurits. 2002. Han. In Norsk Ordbook 'Norwegian Dictionary'.
| Knut Tarald Taraldsen, (Tromsø, CASTL/NORMS): Why that isn't a complementizer |
Kayne has recently argued that English that appears to be used as a complementizer just because it actually is a demonstrative which can also be used as a "relative pronoun", like the West Germanic d-words. From this perspective, the impossibility of using Norwegian det as a complementizer correlates with the impossibility of using it as a "relative pronoun". But why cannot det be used as a relative pronoun? In order to move towards an answer to this, I will explore the internal structure of det and other demonstratives in Norwegian. This investigation will also turn out to shed some light on a second question arising from Kayne's analysis: What is the status of the Norwegian complementizer at?
| Øystein A. Vangsnes (Tromsø, CASTL/NORMS): On the polyfunctionality of WHICH in Övdalian |
The Övdalian lexical item ukin is cognate with English which and Scandinavian (h)vilken, and like its cognates it can be used (i) as a wh-determiner, i.e. to query for a token. But in addition to this ukin has a range of other uses: it has (ii) a predicative use (corresponding to English what ... like), (iii) a adnominal modifier use (corresponding to English what kind of), (iv) a pronominal use in which it queries for human referents (i.e. corresponding to English who), and (v) in its neuter form, ukað, it can be used to introduce yes/no-questions, both matrix and embedded ones. This range of syntactic functions is illustrated by the examples in (1) to (5).
| (1) | Ukin | bil | ir | denn? | (determiner use) | ||||
| WHICH.M.SG.NOM | car | is | yours | ||||||
| 'Which car is yours?' | |||||||||
| (2) | Ukų | sir | å | aut? | (predicative use) | ||||
| WHICH.F.SG.NOM | looks | she | out | ||||||
| 'What does she look like?' | |||||||||
| (3) | Ukan | bil | ar | du? | (modifier use) | ||||
| WHICH.M.SG.NOM | car | have | you | ||||||
| 'What car do you have?' | |||||||||
| (4) | Ukan | al | du | råk | i | Stokkol? | (pronominal use) | ||
| WHICH | shall | you | meet | in | Stockholm | ||||
| 'Who will you meet in Stockholm?' | |||||||||
| (5) | Ukað | ir | du | trät | (eld)? | (question particle use) | |||
| WHICH | are | you | tired | or | |||||
| 'Are you tired?' | |||||||||
| (6) | Ig | wet | it | ukað | å | ir | ungg | eld | gåmål. |
| I | know | not | WHICH | she | is | young | or | old | |
| 'I don't know if she is young or old.' | |||||||||
These matters were investigated during the NORMS fieldwork in älvdalen in 2007. Not unexpectedly, for most of the query types there exist alternative ways of forming them, but although there are some factual uncertainties pertaining to the modifier use of ukin, there is no doubt that this lexical item is quite a polyfunctional element in the grammar of Övdalian. This is a fair statement even though ukin cannot be used as a relativizer, the way English which can, and neither as manner how, the way several adnominal wh-items can in other Scandinavian varieties.
In this talk I will pick up on two issues that arise from the behavior of ukin. First, from a comparative perspective it is interesting to discuss whether the spread of syntactic functions related to ukin follows a certain pattern, i.e. whether the syntactic uses form a conceptual/semantic continuum. Second, I will discuss how to account for the status of ukin within the grammar, and I will basically suggest a solution which draws heavily on the use of (lexical) features.