Usage guidance on what Biber et al. (Reference Biber, Johansson, Leech, Conrad and Finegan1999) call ‘species nouns’ in English (kind/sort/type + of) has a long history, and has tended to concentrate on number concord between a determiner (e.g. this or these) and kind/sort/type in the singular or plural. Thus we have, from four different centuries:
I ought therefore to say this Sort of Goods sells, and not these Sorts of Goods sell. (Baker, Reference Baker1770: 115)
I mean the expression ‘these’ or ‘those kind of things.’ Of course we all see that this is incorrect and indefensible. We ought to say ‘this kind of things,’ ‘that kind of things.’ (Alford, Reference Alford1864: 69–70)
Those kind. ‘Those kind of apples are best’: read, ‘That kind of apples is best.’ It is truly remarkable that many persons who can justly lay claim to the possession of considerable culture use this barbarous combination. (Ayres, Reference Ayres1911: 297)
kind. ‘Those are the kind of numbers that easily solve the mystery … ’ (New York Daily News). Kind and kinds and their antecedents should always enjoy what grammarians call concord. Just as we say ‘this hat’ but ‘those hats’, so the writer above should have said, ‘Those are the kinds of numbers’ or ‘This is the kind of number’. Shakespeare, for what it is worth, didn't always observe the distinction. (Bryson, Reference Bryson2002: 111)
More recently, Keizer (Reference Keizer2007: Chapter 7) has taken a more nuanced approach to what she calls ‘SKT-constructions’ (p. 152):
… these constructions can … be regarded as containing two nominals: a first nominal, N1, which is always one of three lexical items (sort, kind or type), and a second nominal which belongs to an open class. The two are separated by the element of … Both N1 and N2 can occur in the singular and the plural; number agreement between the elements is not required. (2007: 152)
One of the purposes of my research, which aims to analyse a self-compiled corpus of International Academic English to see what determines number marking in these constructions, is to investigate whether modern usage guides, many of which claim to make use of corpus data, reflect or refine older usage guidance, as exemplified above. For example, we find:
these/those sort of. From the 16c. onwards, sort has been used collectively, preceded (illogically) by these or those … Not unexpectedly, the plural form these/those sorts of is also used … The type these/those sort of should now be used only in informal contexts. (Butterfield, Reference Butterfield2015: 763)
*these kind of; *these type of; *these sort of. These illogical forms were not uncommon in the 17th and early 18th centuries, but by the mid-18th they had been stigmatized. Today they brand the speaker or writer as slovenly. (Garner, Reference Garner2016: 906)
but also:
sort of. … [W]hen the phrase is partly or fully pluralized, as in these sort of or these sorts of, it's less clear whether the following noun should be singular or plural. Both constructions are equally well represented in written material from the BNC … (Peters, Reference Peters2004: 507–508)
For my research, in addition to analysing what the corpus authors have actually written, I propose to ask them what they think they should write, and to see how this usage reflects established guidance.
This is where, I hope, readers of English Today will become involved as well. I am inviting you to take part in a survey, which I have posted on the Bridging the Unbridgeable blog here: https://bridgingtheunbridgeable.com/2017/02/14/yet-another-survey/. It is a short survey, starting with an example from Mittins et al.’s (Reference Mittins, Salu, Edminson and Coyne1970) Attitudes to English Usage, just to set a bench-mark. This is followed by 12 examples, all taken from my corpus, and differing from Mittins et al. in that they are all presented in context, typically including the sentence before and the sentence following. Apart from teasing out your attitudes to number concord, the survey also gives you the opportunity to comment at length. I am hoping to use your responses to refine the survey for the authors of my corpus.
Thank you in advance for taking part.
ADRIAN STENTON is a PhD candidate in the project Bridging the Unbridgeable: Linguists, Prescriptivists and the General Public at the Leiden University Centre for Linguistics. For his dissertation, Adrian is conducting a sociolinguistic investigation of usage attitudes to number concord in the species noun phrase in International Academic English. Email: [email protected]
