From In Present-day English, the nouns sort, kind and type (henceforth `SKT-nouns') are well known for syntactically anomalous constructions of various types. Following Keizer's (2002) discussion of the surface pattern D1-N1-of-(D2)-N2 with an SKT-noun in the N1 position,[1] we could distinguish at least binominal and qualifying constructions:

(1) This part of France has many kinds of cheese.
(2) He's a kind of con artist.

These differ in formal properties, discourse properties and semantics. For example, the binominal (1) has N1 as head, while the qualifying (2) has N2 as head. We might also identify another pattern:

(3) These kind of mistakes were happening all the time.

In Keizer's analysis this must be in conventional terms a subtype of the binominal, despite the mismatch of number between the head N1 and both determiner and verb.

SKT-nouns occur in other configurations as well. For example, sort of and kind of can be used before categories other than nouns:

(4) She's kind of scary.

And according to Jennifer Smith (p.c.), younger speakers are using them with increasingly bleached meaning:

(5) ... and I sort of opened the door, and looked out, and I sort of saw Richard ...

All three SKT-nouns can be used in the conventionalised phrase illustrated by

(6) i mean it started their marriage like if he had visitors he used to send her out of the room type of thing

In this usage SKT-of-thing is an adverbial hedge. Now, clear cases of the various patterns can be handled in conventional syntactic approaches, albeit with some difficulty. However, intermediate and indeterminate cases are more awkward. When we bring in the historical dimension there are further issues to consider, among them:

* the different but partially parallel lexical histories of SKT-nouns
* changing membership of the set: kin, manner and others no longer show typical SKT properties, while type is a more recent member than sort and kind
* the appropriate level of diachronic analysis: lexical, syntactic or constructional?
* at what point in the historical development a constructional analysis becomes viable or indeed necessary
* how to recognise and handle transitional phenomena
* whether current change can be idealised out a synchronic account

In this paper I will briefly present an analysis of the present-day facts of SKT-constructions, showing that between the clearly distinct constructions there is categorial and syntactic gradience. I will then sketch the relevant parts of the histories of five SKT-words and the semantic and distributional changes they have undergone, rejecting some of the conclusions in Tabor (1993). Finally I will try to relate the grammatical gradience of present-day English to the semantic histories.

[1] Keizer discusses both determiners and adjectives in the various SKT-constructions.

References

• Aarts, Bas (2000) Modelling gradience. MS., University College London.
• Aarts, Bas (1998) `Binominal Noun Phrases in English'. Transactions of the Philological Society 96: 117-58.
• Denison, David (2001) `Gradience and linguistic change'. In Brinton, Laurel J. (ed.) Historical linguistics 1999: selected papers from the 14th International Conference on Historical Linguistics, Vancouver, 9-13 August 1999. John Benjamins, Amsterdam and Philadelphia PA (Current Issues in Linguistic Theory, 215), pp. 119-44
• Huddleston, Rodney and Geoffrey K. Pullum (2002) The Cambridge grammar of English. Cambridge University Press, Cambridge.
• Hudson, Richard (2000) `*I amn't'. Language 76: 297-323.
• Keizer, Evelien (2002) A classification of sort/kind/type-constructions. MS., University College London.
• Tabor, Whitney (1993) `The gradual development of degree modifier sort of and kind of: a corpus proximity model'. Papers from the th Regional Meeting of the Chicago Linguistic Society 29: 451-65.