2

According to the book, grammatical terms, e.g., subject, object, noun, verb, adjective, etc. should not be defined by meaning, but by grammatical properties.

For example, an adjective has combinations of the following properties:

(1) functions as attributive modifier or predicative complement

(2) can inflect for grade, e.g., better, harder; or form comparative and superlative adjective phrases, e.g., more interesting

(3) can be modified by adverb like "very," or "too"

One thing I notice is that sometimes adjectives are easily identified with these properties because of the context.

For example, the word "hard" in "he is a very hard worker" is obviously an adjective because it is modified by "very" and it functions as an attributive modifier (or, more specifically, heads an adjective phrase that functions as an attributive modifier).

But sometimes adjectives are not easy to identify because the context barely provides anything for you.

For example, in the sentence "The rock is hard" the word "rock" has the only property of an adjective - functioning as predicative complement.

In the book, the authors manipulate the sentence like this to 'test' whether "hard" has any other properties of adjectives. They might add "very" to the sentence to see whether or not "hard" can be modified by "very.

However, sometimes I think they go too far.. Like, in this example from the answer key to the book, the authors try to 'test' whether "barking" can form a comparative phrase by putting it in a completely new sentence (as highlighted).

  1. Which of the underlined words below are adjectives, which are verbs, and which are ambiguous between the two categories in the examples given? Give evidence for your answers.

Example: The dog is barking again.
Answer: *Barking is a verb. It has a corresponding preterite form (The dog barked again), but doesn't have a comparative (*Rex is more barking than Fido). It doesn't accept such modifiers as very (*The dog is very barking). It can't occur in predicative complement function: be in the example is the progressive auxiliary, and can't be replaced by such verbs as become and seem (*The dog seems barking).

Source:http://www.lel.ed.ac.uk/~gpullum/SIEG/exx_ch6.html

So my question is: Are there restrictions on how you can 'test' for grammatical properties? Can I just put the word in question in any sentence to see if it has any other properties?

Improve this question
5
  • 5
    Not just any sentence. You have to have a little expertise before you fool with the grammar rules. Probly you should look at McCawley's first 3 chapters to find out how syntactic tests work.
    – John Lawler
    May 30, 2022 at 23:37
  • 4
    You cannot assign words to parts of speech without any context. The word tiring is an adjective in the sentence The hike was tiring, because you can add very before it. But in the sentence The boy was tiring quickly, it's a verb. And The boy was tiring has two quite different meanings; in one, tiring is an verb and in the other, it's an adjective. Similarly, the word barking is a verb in the sentence The dog is barking, but it's an adverb in the sentence He is barking mad.
    – Peter Shor
    May 31, 2022 at 11:41
  • You might consider asking this at Linguistics where they may have a multi-language perspective (different languages have different parts of speech and differing levels of ambiguity.
    – Mitch
    May 31, 2022 at 13:22
  • Identifying Latinate "parts of speech" is like playing poker with a pinochle deck. Way too many false positives and confusion over rules. It's never necessary to identify one and only one "part of speech" for a word in English; for one thing, the Classic Eight don't list about half of the possible classes because Latin didn't have them (articles, for instance,, which are not adjectives). For another, most of the time, as has been pointed out repeatedly, one can't tell in a given sentence. For a third, nobody cares except teachers. Once you're out of school, you can ignore it.
    – John Lawler
    Jul 25 at 13:49
  • Before one starts applying tests-for-adjectives, one needs to bear in mind (1) that some premodifiers are attributive nouns (football manager; chicken soup ...), (2) some adjectives can only be used in certain slots (some even being necessarily or sometimes postnominal; chicken supreme; responsible) and (3) some adjectives ((a) classifying like nuclear [reaction]; (b) absolute (beware; this term is used in different ways) like dead [sparrow]; (c) otherwise extreme (like enormous [redwood]) are also ungradable (except in quirky usages), failing the second and third tests above.
    – Edwin Ashworth
    Nov 22 at 12:39

1 Answer 1

Reset to default
0

There is at least the following important restriction: you can't rely on grammatical criteria only in order to deduce the grammatical nature of words; there is no absolute test, and in the end you might have to rely on context, that is, on semantic criteria.

Take the following sentence.

  • It is considered as plain country.

"Plain" satisfies all four criteria for adjectives, and could be considered to be the archetypal adjective; yet in a geographical discussion this word can be understood as the noun "plain" that describes a particular sort of relief.

Improve this answer
8
  • 3
    Some words belong in more than one category, and the clauses containing them may be ambiguous. For example, in "I like smoking", the word "smoking" is ambiguous between a verb and a noun.
    – BillJ
    May 31, 2022 at 6:35
  • @BillJ Brilliant deduction!
    – LPH
    May 31, 2022 at 7:04
  • 2
    The point is that we just have to accept that ambiguity exists. In my example, verb is the more salient interpretation (“I like to smoke”). Noun interpretation can be forced by adjectival premodification, as in "occasional smoking". Likewise "Smoking is harmful" -- verb preferred but noun can’t be excluded (cf, "Even occasional social smoking in the pub is prohibited").
    – BillJ
    May 31, 2022 at 7:09
  • @BillJ Therefore the existence of a restriction on the schemes of determination through grammatical critiria.
    – LPH
    May 31, 2022 at 7:18
  • 1
    Yes, but in a geographical context, "it is considered as very plain country" is completely nonsensical. So the text is working.
    – Peter Shor
    May 31, 2022 at 11:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.