English grammar - Roger Berry 2012
David Lee
D2 Count and mass nouns
Section D Extension
David Lee (2001) reprinted from Chapter 8 of Cognitive Linguistics, South Melbourne: Oxford University Press, pp. 137-145.
We have assumed so far that nouns come with their count status already established. Dictionaries tell us that a certain word is count or noncount, or, in certain cases, both. In the paper below, David Lee tries to account for count status using an approach called Cognitive Linguistics, which seeks to find a semantic (i.e. meaning-based) rationale for grammatical phenomena. We have
already seen in C2 that the count status of a noun can change if a specific meaning is needed (e.g. ’beer’ can become ’a beer’). But Lee goes much further, arguing that count status is principally determined by meaning.
While this approach is not free from criticism, the paper certainly provides some interesting insights into the way the count status of a noun may have arisen, historically at least. And it serves as a useful reminder that we should not blindly accept simplified, pedagogic rules based on form without looking for deeper, scientific rules based on meaning (as we saw in A1 regarding the difference between ’some’ and ’any’).
A few words about the text. Lee uses the term ’mass’ rather than ’noncount’, as some other grammarians do. A central concept in the paper is that of ’construal’ (from the verb ’construe’). This refers to how we perceive or make sense of the world. Another important concept (which Lee is disagreeing with) is arbitrariness: the idea that connections (between meaning and grammar in this case) are random or accidental. Finally, note the use of asterisks to indicate unacceptable forms.
8.1 Introduction
It was noted in earlier chapters that one of the defining features of Cognitive Linguistics is its view of the nature of the relationship between form and meaning. The traditional view, firmly asserted by Saussure ([1915] 1974) is that this relationship is characterised by arbitrariness - the forms of language bear no direct relationship to their meaning.
This view is undoubtedly correct in certain respects. For example, there is nothing about the form of words in a particular language that bears any relationship to their meaning - with the minor exception of onomatopoeic words such as crack, crunch, creak, and so on. This does not mean that the form-meaning relationship is always totally arbitrary. In general, the cognitive claim is that grammatical structure is more strongly motivated than has traditionally been thought to be the case. In this chapter, we explore this issue with respect to various types of noun in English.
Nouns can be divided into a number of different subclasses with respect to their inflectional properties.
□ Class A These are prototypical nouns, having both a singular form and a plural form: cat ~ cats.
□ Class B These nouns have only a singular form: equipment ~ *equipments; furniture ~ *furnitures; crockery ~ *crockeries.
□ Class C These nouns have only a plural form: *scissor ~ scissors; *trouser ~ trousers; *clothe ~ clothes.
□ Class D These nouns have both a singular and a plural form but they are identical: sheep ~ sheep; deer ~ deer; salmon ~ salmon.
□ The distinction between class A nouns and class B nouns has a number of other grammatical reflexes.
□ The singular form of class B nouns occurs without a determiner in positions typically occupied by noun phrases (Furniture is useful, I bought furniture,
I’m looking for furniture), but this is not generally true of class A nouns (*Cat is useful, *I bought cat, *I’m looking for cat).
□ The indefinite determiner a occurs with class A nouns (a cat) but not with class B nouns (*a furniture).
□ The determiner much occurs with class B nouns (much furniture) but not with class A nouns (*much cat).
□ Expressions such as a lot of occur with the singular form of class B nouns (a lot of furniture) but not with the singular form of class A nouns (*a lot of cat).
These observations have led linguists to make a terminological distinction between ’count’ nouns (class A) and ’mass’ nouns (class B). The question on which I will focus in this chapter is whether this distinction is motivated or arbitrary.
Certain examples seem to support the view that it is arbitrary. For example, there seems to be no obvious reason why vegetable is a count noun but fruit is (normally) a mass noun. Ware (1979: 22) makes this point in the following terms:
Turning now to why it is that words sometimes have count occurrences and sometimes mass occurrences, we are immediately faced with the problem of a tremendous amount of variation that appears unnecessary and inexplicable (. . .) There is a count-mass difference between fruit and vegetable but they apply to things that for all accounts and purposes seem to be alike. Nor can I see anything that would explain the count-mass difference between footwear and shoe, clothing and clothes, shit and turd or fuzz and cop.
Other contrasts that could be taken to support the arbitrariness view involve examples such as noodle (count) and rice (mass), bean (count) and spaghetti (mass). For example, why do we refer to lots of noodles in a bowl as these noodles (using the plural form of a count noun) but to lots of grains of rice in a bowl as this rice (using the singular form of a mass noun)?
In many cases, however, there is an obvious basis for the distinction between count nouns and mass nouns. There is a strong tendency for count nouns to refer to ’objects’ and for mass nouns to refer to ’substances’. For example, the fact that cup, cat, and table are count nouns whereas custard, water, and sand are mass nouns seems far from arbitrary. What then is the distinction between ’objects’ and ’substances’, and can it be used to motivate the count/mass distinction in general?
8.2 Count and mass phenomena
Let us take as the starting point the fact that solid physical objects such as bicycles and cats are typically designated by count nouns, whereas liquids such as water and oil are typically designated by mass nouns. One salient difference between solid objects such as bicycles and cats on the one hand and liquids such as water and oil on the other is that the former have a characteristic shape and well-defined boundaries, whereas the latter lack such a characteristic profile, moulding themselves to the shape of their container. And whereas solid objects have an internal structure typically consisting of discrete components (a bicycle has a frame, wheels, handlebars, pedals, and so on), liquids tend to be internally homogeneous. One consequence of this is that any particular ’segment’ of a liquid counts as equivalent to any other segment. For example, if I dip a cup into a pond and remove some water, then this particular segment of ’water’ is, for all practical purposes, identical to any other segment of water that I might have scooped up. But most parts of a bicycle are different from the other parts.
Consider now such phenomena as slime, mud, and silt. In terms of their texture, these are intermediate between solids and liquids but they resemble liquids more than solids with respect to their external boundaries (they lack a characteristic shape) and internal structure (they tend to be homogeneous). It is therefore not surprising that the corresponding nouns belong to the mass category (*several slimes, *these muds).
The same applies to internally homogeneous solids such as earth, clay, and cement. Chemically speaking, of course, all these substances consist of particles (as indeed does water), so that in this sense they are made up of large numbers of elements. But this is irrelevant to practical human concerns - and therefore to language.
When we come to consider phenomena such as sand, sugar, rice, soot, dust, and so on, we begin to approach the point where there is a potential motivation for construal in terms of a collection of individuated objects rather than a substance. The particles of which sugar and sand are composed are at least discernible to the human eye. In principle, therefore, there is no intrinsic reason why English should not have a word that refers to a single grain of sugar (for example, flig), in which case I might point to a pile of sugar and say Here are some fligs. But there are obvious reasons why there is no such word. Whenever sugar manifests itself to us, it always does so in the form of a conglomerate of thousands of ’fligs’, tightly packed together, so that the word flig would serve little practical purpose. Moreover, sugar in this form behaves just like water. It moulds itself to the shape of a container and is internally homogeneous.
Some of these points also apply to noodles, but grammatically speaking we have now crossed the mass/count frontier, since noodle is a count noun. Individual noodles are bigger than individual grains of rice, which provides some motivation for the grammatical distinction. (It is easier to eat a single noodle than a single grain of rice.) It has to be said, however, that a single noodle is unlikely to be of great interest to anyone. To take a similar example, there is very little difference between the size of the particles that make up a pile of gravel and those that make up a pile of pebbles, but gravel is a mass noun, whereas pebble is a count noun.
Although there is undoubtedly a certain degree of arbitrariness in these cases, located as they are at the boundary between objects and substances, this does not mean that there is no semantic motivation for the count/mass distinction in general. In fact, the indeterminacy that we find in cases such as gravel and pebbles is precisely what one would expect in a theory in which the objective properties of entities are subject to processes of perception and construal. Time and again in language, we come across situations where the distinction between two categories is semantically motivated, but where the behaviour of phenomena located at or near the boundary is not wholly predictable. Just as a pile of gravel is not a prototypical mass phenomenon (since it is composed of a number of perceptually distinguishable particles), so pebbles and noodles are not prototypical count phenomena, given that individual pebbles and noodles are generally not as perceptually salient as individual cats and bicycles and are of much less interest to human beings.
One strong argument for the claim that there is a semantic basis for the count/ mass distinction is the fact that some nouns have both count and mass uses, associated with a clear difference in meaning. Consider:
1. (a) Could I have a potato? (count)
(b) Could I have some potato? (mass)
2. (a) I’ll have an egg. (count)
(b) I’ll have some egg. (mass)
3. (a) I’d like a pumpkin. (count)
(b) I’d like some pumpkin. (mass)
4. (a) There were a lot of newspapers in the box. (count)
(b) There was a lot of newspaper in the box. (mass)
5. (a) There’s a glass on the table. (count)
(b) It’s made of glass. (mass)
Potatoes, eggs, and pumpkins normally manifest themselves as unitary, individuated, countable objects and they may retain this character in cooking and serving. However, if a potato or pumpkin is mashed or an egg scrambled, its character changes. It becomes a homogeneous substance from which portions can be removed or further portions added without changing its character. These examples show that, strictly speaking, the terms ’count’ and ’mass’ do not refer to types of noun but to particular uses of nouns, though it is true that many nouns normally appear only in one or the other use-type, given the nature of the entity that they designate.
It would be a mistake, however, to attempt to motivate the count/mass distinction purely in terms of the physical properties of phenomena. Consider, for example, the case of liquid substances. I began this discussion by observing that liquids are typically designated by mass nouns (I’ll have some water, There’s beer in the fridge, He drank a glass of wine). However, count uses of these nouns are by no means unusual.
6. There were several wines on show.
7. He drank a few beers.
8. The waters were rising.
These uses have a variety of motivations. As far as wine is concerned, human beings find it highly relevant to their everyday concerns to divide the phenomenon into various subtypes. Since each such subtype is an individuated entity, it is designated by a count noun, as in (6). Example (7) can also be interpreted in this way (that is, as meaning that he drank a few types of beer), but it is more likely to be used to refer to a rather different kind of countable phenomenon such as the contents of a container (He drank several wines, Two sugars please). In (8) floods are typically fed by water from different sources (for example, different rivers), so that even after they have merged, they can still be conceptualised as different entities. (This usage may also be motivated by the fact that flood water manifests itself in different places.) In objective terms the nature of the phenomenon in (8) is of course no different from any other manifestation of water as a mass. If the claim were that objective criteria determined linguistic form, then this usage would constitute a powerful counterexample. But if language reflects conceptualisation, it is not difficult to identify a cognitive basis for the example.
So far, it has been argued that mass phenomena are characterised by internal homogeneity. Let me now consider nouns such as cutlery, furniture, and crockery. These phenomena seem to constitute a counterexample to the argument, since they refer to collections of discrete, countable entities. The motivation for their assimilation to the mass category has to do with the level at which the concept applies. A set of knives, forks, and spoons can either be construed as a collection of separate objects performing different functions (cutting food, picking up food, stirring liquids) or as a collection of objects which manifest themselves contiguously and which all perform the same function (facilitating the consumption of food). At this level, any part of the phenomenon counts as equivalent to any other part. Similarly, a collection of chairs, tables, and cupboards is subject to alternative construals. We can think of them either as a group of separate objects or as a unitary entity with a single function - that is, as ’furniture’.
Note, however, that more general levels of categorisation do not always produce a mass noun in English. The concept ’tool’, for example, constitutes a superordinate category with respect to hammers, screwdrivers, drills, and so on, but tool is nevertheless a count noun. As in the case of ’cutlery’, we are dealing here with an experientially related set of entities that perform different functions at one level and a single function at a more abstract level. But the grammatical character of the form tools continues to foreground the essentially plural nature of the phenomenon. On the other hand, the same set of entities could be designated by the mass noun equipment, which foregrounds their functional unity.
This observation helps to explain the contrast between fruit and vegetables. Like cutlery, furniture, crockery, and equipment the word fruit is a manifestation of a general pattern in the language, such that the grammatical character of the word foregrounds the pragmatic contiguity and functional similarity of the entities that constitute the category. Vegetables is a manifestation of a different pattern, whereby the abstraction to a superordinate level is realised lexically but where the grammatical character of the word continues to highlight the essentially plural and diverse nature of the phenomenon, as in the case of the word tools. Again, there is nothing in reality that requires the language to work in this way (that is, nothing that requires a grammatical distinction between fruit and vegetables), but there is no difficulty in identifying characteristics of the phenomenon that motivate the distinction.
The contrast between clothing and clothes constitutes a similar case. Whereas the unitary nature of the phenomenon is captured by the fact that the word clothing has only a singular form, the diverse nature of the objects that constitute the category and the fact that different items normally occur together are reflected in the fact that clothes has only a plural form.
8.3 Nouns lacking a singular form
So far, we have focused on the distinction between class A nouns (count) and class B nouns (mass). However, it was noted at the beginning of the chapter that there is also a subcategory of nouns (class C) that lacks a singular form, clothes being a member thereof. Nouns of this type can be further divided into a number of subclasses according to the nature of the motivation.
As far as ailments such as measles, mumps, shingles, hives, and haemorrhoids are concerned, the motivation for the inherently plural character of the corresponding nouns is obvious. This is also true of words such as pants, braces, scissors, shears, binoculars, tweezers, clippers, tongs, goggles, spectacles, glasses, and so on, though the motivation for the inherent plurality of these forms has not always been recognised. Gleason (1961: 224), for example, writes: . . . by a convention of English, pants is plural. Interestingly enough, this is not an isolated case; compare trousers, breeches, shorts, slacks, etc. This whole group of words are grammatically plural with no evident semantic justification.
It is interesting to speculate about the factors that led Gleason to make this surprising observation, since it cannot have escaped his notice that a prominent feature of these objects is that they consist of two identical parts. Gleason’s comment presumably derives from the fact that a pair of trousers is a unitary object, so that in one sense there is no obvious reason why the language should treat such a garment differently from any other unitary object, particularly when other items of clothing such as coats and shirts, which also possess two identical parts, are designated by count nouns. Certainly there are many languages that use ordinary count nouns to refer to a pair of trousers (pantalon in French, Hose in German). Again, however, this clearly does not mean that the use of a plural noun is unmotivated, nor that the distinction between trousers on the one hand and coat or shirt on the other is arbitrary. After all, the sleeves of coats and shirts do not constitute as great a proportion of the whole garment as do the legs of a pair of trousers. What is missing, then, from Gleason’s way of thinking about these examples are the notions of foregrounding, construal, and motivation.
8.4 Nouns with identical singular and plural forms
The class D words identified at the beginning of this chapter (salmon, cod, sheep, pheasant, mackerel, and so on) differ from other noun classes in that they have identical singular and plural forms (compare this salmon~ these salmon, one cod ~ several cod, and so on). The fact that most of the phenomena in question belong to the same conceptual domain suggests that there is some underlying rationale here - that the class of nouns with this unusual grammatical property do not constitute an arbitrary set.
The semantic property shared by most of these nouns is that they traditionally belong to the domain of hunting and fishing. In other words, the phenomena in question constitute a food resource. When someone catches a fish, it is both an individual entity and a representative of the species to which it belongs. This latter property is salient in this context because it involves characteristics that are crucial to the general fishing scenario - whether the catch is edible, how it will taste, how many will be needed to make it a viable meal, and so on. To put it slightly differently, when you are fishing for ’salmon’ (or indeed simply buying ’salmon’ at the fishmonger’s), it is relatively immaterial which particular individual you acquire. What is important is that it is ’salmon’ rather than ’cod’ or ’mackerel’. This property relates entities of this kind to mass phenomena. Just as any arbitrary portion of water is equivalent to any other portion, so any individual salmon is as good a representative of its species as any other from the point of view of the consumer. The grammatical character of these nouns, therefore, seems to be a reflex of a general ambivalence concerning the individuated and mass aspects of the phenomenon. Their individuated character is often highly salient (the difference between catching one salmon and several can be important), so that it is useful to have a singular ~ plural contrast, but the absence of explicit plural marking seems to be a reflex of the fact that the individual is an arbitrary manifestation of a general resource.
[. . .]
8.5 Conclusion
The discussion in this chapter supports the general claim of cognitive linguists that grammar is a much less arbitrary phenomenon than has traditionally been thought to be the case. It is important not to overstate this claim. Cognitivists do not argue that grammatical properties are invariably explicable in terms of underlying cognitive or pragmatic factors - that is, that grammar is wholly determined by such factors. As has been noted, the fact that vegetable is a count noun whereas fruit is normally used as a mass noun is an arbitrary fact to some extent. Given that both terms involve a generalisation across a somewhat disparate set of phenomena, the option is available for the concept to be treated either as a collection of individuated entities or as a mass. This does not mean, however, that the grammatical count/mass distinction is entirely arbitrary. The crucial concept that allows a degree of arbitrariness to be reconciled with the notion of motivation is construal. Since this concept specifically allows for alternative ways of conceptualising a situation, it allows for phenomena that are perceptually similar to be treated either as grammatically similar or as grammatically different.
Certainly we sacrifice here some of the determinacy that linguists have often hankered after in their search for linguistic generalisations, but given the fact that language is a social phenomenon subject to the constraints of convention and cultural variability, total determinacy may prove to be an unrealistic goal in linguistic description and theory. In the next chapter we pursue this theme with respect to certain grammatical characteristics of verbs, taking up the count/mass distinction in a more abstract form.
Questions, suggestions and issues to consider
1. To what extent do you agree with Lee’s hypothesis? Do you feel that the way you ’construe’ nouns conditions the way you use them grammatically? Or do you simply follow usage: what other people say, or what teachers and grammars tell you to say?
2. Take some of the nouns he discusses and think about how you construe them. For example, do you feel that the noncount status of words like furniture and information is illogical? In other words, do you construe them as individual, divisible entities? If so, would this allow you to use them as count nouns, in contravention of the ’rules’ of English grammar?
3. Lee does not discuss collective nouns such as government or team (see A2). How would they fit into his four classes at the start? Do they provide evidence for or against his theory?
4. Could this approach account for the nouns described in C2, such as reason or paper, where the count/noncount difference is associated with an unpredictable difference in meaning?
5. Lee claims that the perception of meaning (construal) conditions the grammar of nouns. But is it not also possible that the grammar conditions the meaning, i.e. there is a two-way relationship? What if we encounter a word first, before we have any idea of what it is referring to? Might the fact that it is noncount lead us to think of it as a mass, regardless of its true nature?
6. One of the criticisms of this approach to nouns is that equivalent words in two languages may have different count status. For example, information is noncount in English but count in French (where it is spelt exactly the same), and the French word for furniture (noncount) is meubles (count). Does this mean that French and English people regard ’information’ and ’furniture’ differently? What about when bilingual speakers switch from English to French? Do they suddenly change their perception of the world? Or is this a grammatical accident that is conditioned by the language? To what extent do you feel grammatical phenomena are as they are because of convention, or repetition, rather than creativity?