A facebook exchange back on the 6th, between Andrew Carnie (professor of linguistics and dean of the Graduate College at the Univ. of Arizona) and Karen Chung (associate professor at National Taiwan University, teaching courses on linguistics and English).

Andrew: [Student], who only came to class less than 50% of the time, and turned in a bunch of assignments (really) late: These homeworks are way. too. hard. It’s unfair.

Karen: “Homework” as a countable noun? Is he/she a native speaker of English?

Academics will recognize Andrew’s note as the plangent lament of a professor facing the grading tasks at the end of a term, confronted with a self-entitled student who believes they are really smart, so preparation outside of class shouldn’t take much work (and they should be able to ace the final without much studying).

But what Karen picks up on is the use the noun homework as a C(ount) noun, clearly so because it occurs in the plural form homeworks here; for the M(ass) noun homework, the usage would be: This homework is way. too. hard. Or else: These homework assignments are way. too. hard.

Much as I sympathize deeply with Andrew’s lament — having had nearly 50 years of similar experiences (fortunately far outweighed by students who were a delight to teach) — what this posting is about is the C/M thing. There’s a fair amount to get clear about first, and then I’ll have some analysis, some data, and some reflections on larger matters (language use in particular communities of practice, the tension between brevity and clarity as factors in language use).

Background: C/M. There’s a Page on this blog of my postings on the distinction between the morphosyntactic categories C and M, with a special place given to my detailed summary of the topic: the handout for my 2001 SemFest talk on “Counting Chad”, on the C/M distinction in English, with notes on chade-mail/email, and ice plant; on the grammatical categories C, M, SG, PL, E, and I; and on a default association between this morphosyntactic classification and semantics — taCsiM (things are C, stuff is M).

But the world is full of entities that present themselves to us in ways that, depending on the nature of our interests in them, we can view as either an assortment of individuals (things) or an overall agglomeration (stuff). Woody plants that are relatively small and often grow (naturally, or by our design) in stands can be viewed as things — referred to by PL C nouns like bushes or shrubs — or as stuff — referred to by SG M nouns like shrubbery.

When things views and stuff views are both available, a language might have a M lexical item for the stuff view and a syntactic construction built on that for the things view: the English M items luggage and furniture, with corresponding C piece of luggage, piece of furniture.

Or it might have both M and C uses for the same formal base, as when English ice plant (referring to a number of groundcover plants with fleshy leaves) is used either as M (There’s a lot of ice plant growing on that dune) or C (There are a lot of ice plants growing on that dune).

The homework case. Karen Chung’s first reaction to the novel (to her) PL C nominal these homeworks was to suppose that the speaker was not a native speaker of English, but of some language in which the usual lexical item for reference to assignments for student work outside of class was PL C (like English homework assignments) rather than SG M (like English homework). That’s an entirely plausible idea, and such unfortunate loan translations do occur.

But Andrew Carnie’s student was unquestionably a native speaker. In fact, both Andrew and I found this use of homeworks entirely normal, and noted that it’s common in American college usage, by both instructors and students, and has been for decades. (So it’s probably significant that Karen’s been off in Taiwan for some time.)

Now, at this point, some people become alarmed that the appearance of PL C homeworks means that it has replaced SG M homework, that somehow homework has become homeworks (so that their homework is no longer acceptable). This is not at all what’s happened. All that’s happened is that some speakers now have two lexical items homework: a M noun in examples like I hate doing homework (which, so far as I know, is uncomplicatedly acceptable for everyone), plus a specialized C noun with the meaning ‘homework assignment’.

The situation with homework is much like the situation with e-mail/email, in fact. C e-mail (as in There are a lot of e-mails in my queue) has not supplanted the earlier M e-mail (There is a lot of e-mail in my queue) — an item that’s predictably M because its head, mail, is M — but is merely a different lexical item, with a somewhat different meaning. The C noun is of course the only one available in contexts where actual counting of messages is involved (There are 26 e-mails/*e-mail in my queue; to use M e-mail, you must embed it in a larger construction: 26 e-mail messages, 26 pieces of e-mail), and the M noun is the only one available in contexts that select reference to stuff (My e-mail expands every week; cf. The number of my e-mails expands every week).

Both for homework and for e-mail, the difference between the M and the C nouns is not just a matter of viewpoint (thinking of the very same referent as stuff or as things), but involves a semantic distinction that follows from the difference between things and stuff.

Too much/many. From Mr. Dub’s Language Lab site on 11/3/15:


“Teaching: too many, too much, & enough”: the scale not enough – enough – too much/many and its accompanying syntax

Imagine a 15-week course that meets three times a week. As the teacher, you you assign homework due at every course meeting (with some allowance for mid-term exams, so that there will be, say 35-40 homework assignments over the term). Each assignment is small, but you believe that it’s important that the students constantly engage actively with the material of the course and get feedback from you on that engagement, so that they can improve — since much of the the material is cumulative, and the later stuff can’t be mastered unless the earlier stuff has been, while at the same time, the course might involve a number of separate sub-units, each of which needs to be mastered on its own (in intro linguistics, say, everything from phonetics and phonology to semantics and pragmatics).

A student taking this course might complain that there are too many homeworks; they’re doing homework all the time. That’s the point, of course, but many students would prefer to put their effort into a small number of all-out work binges. (Oddly, no one seems to do this in learning physical or practical skills — no all-nighters in honing your skills at basketball or carpentry. But let that pass.) For these students, the very same tasks packaged into about 10 or so homework assignments would be fine, just enough homeworks. Also, let us grant, just enough homework, whether packaged into 35-40 assignments or just 10 or so.

Now consider a somewhat different course arrangement, well suited to the student who prefers fewer but meatier assignments. Imagine a course in modern English syntax that runs for a 15-week term, but with only 10 homeworks, each of the form:

read chapters x and y in The Cambridge Grammar of the English Language and write a ca. 15 page critique of the treatment there of one of the following four topics…

(There are 20 chapters in CGEL; the central ones are about 100 pages each, of dense text with illustrative examples.) I have taken literature courses  arranged much like this, but of course with more pages to read per week.

A student might complain about this course that while there are not too many homeworks, there is in fact too much homework. Each homework isa real bear.

The point of all this is that though C homeworks and M homework are crudely similar in meaning, they are in fact distinct, in just the way you’d expect from the difference between things and stuff. This is true despite the fact that in many contexts the distinction is inconsequential.

Beheading. An analysis that goes some way in accounting for the difference in meaning for C homework (one homework, many homeworks) vs. M homework (much homework) and for C e-mail (one e-mail, many e-mails) vs. M e-mail (much e-mail) is that in both cases the C variants are beheaded nominals — of homework assignment and e-mail message, respectively.

As such, they inherit the morphosyntactic properties of the missing head: assignment is C, so beheaded homework is too, even though the modifier homework is M; and similarly for e-mail. (There’s an inventory of my postings on beheading here.)

Intriguingly, OED3 (Sept. 2011) on the noun homework provides a definition that would be consistent with C uses of the noun —

2. Schoolwork assigned to a pupil to be done outside lesson time (typically at home). In extended use: an assignment or exercise to be completed in one’s own time.

but the cites, from 1852 on, are all M.

Communities of practice. Some examples of C homework pulled up in the first few pages of a Google search:

Dartmouth College: ENGS 250: Turbulence in Fluids  Syllabus & Schedule – Fall 2015
Weekly graded homeworks (Thursday to Thursday) (link)

GitHub ScPo-CompEcon syllabus 2018
Workflow of Homeworks. Requirements for Homeworks: I invite you to do the homeworks in teams of 2. … You have two weeks for each homework.(link)

Univ. of  Ill. at Chicago CS 361: Computer Systems
Homeworks will consist of approximately six programming projects with duration between one and two weeks. Homeworks will center around investigating or creating C programs. Be sure to consult the online handout, the TAs, or the professor if you have any questions.(link)

Typically, uses in connection with computer science or engineering courses, in the US. The usage has spread to other academic fields (Andrew Carnie and I know it in linguistics back a few decades), and, I think to use in high schools; beyond the US, I don’t know, but this would be easy to investigate.

In fact, the first cites that Google Ngram pulls up in books are from 1982, 1983, and 1990, all in computer or engineering books. So it seems the usage began in this rather narrow community of practice around 40 years ago and spread from there.

It’s not uncommon for beheadings to originate in specialized communities of practice, serving as group-talk, abbreviated jargon within a group that will understand the usage: the C noun attending for attending physician in medical communities, for example.

Brevity and clarity. Brevity is obviously a virtue within relatively small communities, where it allows for speedier communication (and also helps to   foster group coherence). As always, though, brevity is bought at the cost of clarity. In this case, at the cost of ambiguity — ambiguity in SG homework, which can be either C or M. The NP this homework, and a number of others, is ambiguous between C and M, but the distinction between the two readings of This homework is too hard or I can’t do the homework is too subtle to be of consequence in any context I can imagine. So the ambiguity is no bar to the spread of the C usage (similarly for e-mail).

And now we have homeworks and e-mails.



Leave a Reply

%d bloggers like this: