How did grammarians determine that the Present Continuous is an aspect?

Question

The three variants of the present tense are:

• [X] sits (Simple)
• [X] does sit (Emphatic)
• [X] is sitting (Continuous [also called Progressive])

This is something that I was taught in school at such an early age that I have become used to seeing and thinking of the form as a single unit. Is sitting is just the present continuous of to sit. So it wasn't until very recently that it occurred to me (insert Philosoraptor meme :D) that, in a sentence like "Jane is sitting":

This construction has the exact same form as [Subject] - [Linking Verb] - [Predicate Adjective]! I find that interesting.

Since that's not the normal way of interpreting such a sentence, I'll provide examples by way of analogy to other predicate adjective constructions.

---

1 - Sitting used as an attributive adjective

For context, on its own (when it isn't being used to form a continuous tense), the present participle is used as an adjective (its use as an adverb is much less common). Either pre-nominally ("a swimming fish"), post-nominally ("a fish swimming"), or to govern a phrase which functions as an adjective ("a fish swimming rapidly upstream"). Here are two examples where "sitting" is used as an attributive adjective: "a sitting duck"; "the sitting President." You could substitute a regular attributive adjective in the exact same context: "a large duck"; "the current President."

2 - Changing an attributive adjective to a predicate adjective

When one of those same regular adjectives follows a form of the linking verb to be, it becomes a predicate adjective (instead of attributive). Examples: "The geese are large"; "this schedule is current."

3 - Why not say sitting is used as a predicate adjective?

Since any adjective can function either attributively or as a predicate, I'm curious as to why the sentence "Jane is sitting" can't be analyzed as [Jane (Subj.)] [is (copula)] [sitting (Pred. Adj.)]? If someone asked me that question (why can't it be?), I would say it's because is + sitting automatically "coalesces" into the Present Continuous form when those two words occur together. So, just like string concatenation takes precedence over numerical addition in JavaScript (where 2 + "2" is "22" and not 4), the present continuous form takes precedence over the predicate adjective construction, which precludes the latter's validity. Just like if you actually wanted 4 from the JS expression, you would have to explicitly type-cast it as 2 + Number("2"), because it would never be interpreted that way on its own.

But, since [S] - [c] - [PA] seems to be a logical breakdown of the sentence, and the meaning wouldn't substantially change if the sentence is interpreted that way, how did the designation of present continuous tense come about? I'm interested because normally, it seems to violate Occam's razor ("don't assume more complexity than you have to without a good reason") to analyze something that could come under a general case as belonging to a special case instead. But since it is a special case, reading the form is/are [X]ing as the present continuous of to [X] is so automatic that it seems weird to think of [X]ing as functioning as a predicate adjective instead. So how did that classification come about originally?

I'm asking this here because resources on the present continuous and on the topic of aspect go into a lot of detail about the usage of those terms, but not so much on the history of how they came to be classified the way they are. I have seen a lot of good history/etymology-oriented answers on this site, so I hope this question isn't off-topic.

Answer 1

So, the thrust of the question seems to be why we don't regard 'sitting' in 'Jane was sitting' as an adjective, and why is 'is sitting' considered a present continuous construction there, as opposed to a copular predicate adjective construction.

This is one of those times when it becomes very important to distinguish between three very different linguisticky things:

• meaning
• grammatical relations
• word categories

Let's consider an example similar to the Original Poster's:

1. Jane is entertaining.

This sentence is ambiguous between a reading where 'entertaining' is a verb and one where it's an adjective. The verbal reading is, further, ambiguous between one where Jane has guests right now and another where she is providing amusement or enjoyment to some being(s) or other. In order to keep the meaning of the verb and the adjective as similar as possible, let's just consider the latter case where the verbal reading means that Jane is amusing someone, and where who exactly is recoverable from the context.

Next notice that although the verb entertain has a dynamic meaning, the adjective does not. The verbal reading of 'Jane is entertaining' also means that Jane is entertaining now. Jane cannot be asleep in the verbal reading of 'Jane is entertaining' (i.e. 'Jane is entertaining [some people]'), but she obviously can in the adjectival one (i.e. 'Jane has this quality where people would find her amusing to be with'). In fact, in the sentence where 'entertaining' is an adjective, Jane could have been holed up in solitary confinement for the last decade, and currently be holed up still, unseen by anyone. While verbs can have stative or dynamic meanings (where they describe either states or actions), adjectives have stative meanings - even when they describe a propensity to act in a certain way, like the adjectives explosive or violent or loquacious. Those adjectives describe propensities or tendencies not actions themselves.

We must come to the conclusion, therefore, that there is a semantic difference between the construction where entertaining is an adjective and the one where it is a verb.

The Original Poster discusses cases where they believe that 'sitting' is an adjective, for example when it occurs as an attributive modifier of a noun. In fact, occurring as an attributive modifier of a noun is a grammatical relation not monopolised by adjectives, but shared by verbs as well as nouns themselves and even determiners.

Consider the following example:

• the entertaining husband

This could mean either a husband who generally has the propensity to amuse (adjective), or a husband who is in the process of entertaining people or hosting people (verb):

• the entertaining husband should provide canapés

Similarly, the verb 'sitting' describes people who are currently in a state of sittingness, or who currently carry out the action of sitting: 'the sitting president' or 'a sitting duck'.

However, meaning is a tricky and often unreliable way to distinguish parts of speech, or categories of phrase (notice that OP's original reasoning seems to have a semantic element, in that they argue that the verb sitting describes the subject Jane in the same way that nice would in a copula construction). A much better way is to carry out syntactic tests. Some of these have been demonstrated by @BillJ in his answer here.

We can complement BillJ's tests with some more. Verbs sometimes take direct objects, in other words they can take noun phrases (NPs) as complements. Only a couple of adjectives, one of which is the adjective worth, can take an NP complement. So in the following sentence entertaining must be a verb, not an adjective:

• Jane was entertaining the audience.

Also many adjectives can be pre-modified by the comparative adverb more, whereas verbs cannot:

• Jane was more entertaining than Bob.
• *Jane more entertains than Bob. (ungrammatical)

---

I'd like to move the discussion up a gear here. So let's distinguish three types of construction with the auxiliary verb be (we'll be using auxiliary verb to refer to a class of verbs, not an incidence of a case of be being followed by another verb):

• be dismantled (passives)
• be arriving (continuous constructions)
• be nice (copular constructions)

Let's call the different bes here passive be and continuous be and copular be respectively.

So, one way that we can show the copular and continuous constructions to be different is that a copular be verb phrase can function as the complement of a continuous be:

• Jane was being entertaining.

This can't work the other way round:

• *Jane was entertaining be being. (ungrammatical)

So, it seems that this is clear evidence that the two constructions are not equivalent. One can be embedded as part of the other but not vice versa.

Lastly, why consider the [be + present participle] construction an aspect?

An aspect is usually an inflection on a verb which indicates how the state or action it denotes is being represented/viewed in time. Clearly the construction under consideration is not an inflection of a verb. However, we can also consider so-called periphrastic aspect, which is when a multi-word verbal construction is used to convey the same kind of meaning. This is similar to the concept of periphrastic tense. Tense is usually considered an inflection on a verb which characteristically is used to refer to present, past or future time. However, some linguists conceive of the perfect construction in English as a periphrastic past tense. Or the use of will + verb as a periphrastic future. [I don't subscribe to these ideas about periphrastic tense in English, myself. That doesn't count for much.]

Well, it does indeed seem reasonable to consider the [be + present participle] construction a periphrastic aspect, where a combination of the auxiliary be and a present participle -ing suffix on the lexical verb serve to indicate that the state or action denoted by the verb can be considered as being concurrent with another time (which can cause all kinds of implicatures - for example that this state or action is/was/will be only temporary).

It is further reasonable to conceive of the construction in this way because it can combine with all kinds of other verbal constructions: ones with modal verbs, present tense auxiliaries, past tense ones, perfect constructions and so forth. And in each of these cases, the aspectual meaning conveyed by the [be + present participle] is the same. And this is the same kind of job that is done by aspectual inflections on verbs in other languages.

And of course, this meaning and the meaning conveyed by a simple copular construction are utterly different. In fact in a copular construction, the presence of the verb be adds nothing to the meaning really apart from carrying the features of tense and person.

Conclusion

It is a worthy speculation, then, that the -ing form in a present continuous construction might be a predicate adjective. However, the result of any rigorous investigation will show that this is not the case. Is the present continuous a periphrastic aspect? Well, no. But it's reasonable to consider the continuous construction [be + present participle] a periphrastic aspect. But if it is, like aspects that occur as verbal inflections, it combines with different tenses and modal constructions.

Answer 2

Regarding the first point in your question:

Jane is sitting/nice.

"Sitting" differs from "nice" in that:

(a) It can't be modified by "very", which can modify adjectives: We can't say *Jane is very sitting.

(b) It can't occur as complement to complex-intransitive verbs like "become": We can't say *Jane became quite sitting.

(c) It can't occur as complement to complex-transitive verbs like "find": we can't say *I found Jane quite sitting.

Compare Jane is/seemed very nice; I find Jane very nice.

"Nice" has the properties of indisputable adjectives and hence must belong in that class, but "sitting" does't have the distinctive properties of adjectives and hence can only be a verb.

Regarding the second point in your question, have you looked in a modern scholarly text book?

Answer 3

In addition to BillJ's excellent examples of how is sitting differs from is nice, it's also worth considering some ways that is sitting is similar to sits.

The biggest is that present participles accept the same complementation patterns that their verbs always do; for example:

• direct object: "I ate dinner" → "I'm eating dinner"
• indirect object + direct object: "I told her the truth" → "I'm telling her the truth"
• subject complement: "I grew older" → "I'm growing older"
• direct object + object complement: "I found it difficult" → "I'm finding it difficult"
• direct object + infinitive: "I asked him to stay" → "I'm asking him to stay"
• direct object + adverbial: "I treated them well" → "I'm treating them well"

Notably, in cases like "surprising" where a participial adjective has developed from a participle (which happens a lot: English has many such adjectives), the complementation patterns often distinguish the two uses; contrast "She was surprising us almost every day with some new skill she had learned" (participle, verb form, takes direct object) with "Her sudden interest was surprising to us, and we did what we could to encourage it" (adjective, requires preposition). (Admittedly, English does have a few adjectives that can take a noun phrase complement with no preposition — "It's worth a lot", "It's unlike anything I've ever seen" — but they are a rare exception to an almost universal pattern.)

Additionally, present participles participate in the same idioms that their verbs generally do:

• "I sat up" → "I'm sitting up"
• "I fell head over heels" → "I'm falling head over heels"
• "I lost my head" → "I'm losing my head"

---

That being said, you may notice that I said that present participles take these complements and participate in these idioms, rather than that present continuous verb forms do. That's because these participles can do the same in other constructions where they appear: "I began eating dinner", "I remember telling her the truth", "I look forward to growing older", "I was afraid of finding it difficult", "I don't regret asking him to stay", "I'm proud of treating them well", "I saw her sitting up", "I enjoyed falling head over heels", "I admit losing my head". (Note: traditional grammar would count some of these examples as "gerunds" rather than "participles", but the result is the same either way.)

So this still leaves the question of why grammarians would treat "is sitting" as a verb form, or at least, as a distinct construction meriting its own name; neither BillJ's examples nor mine really justify that.

I will tentatively offer three reasons that this might be:

1. Grammarians are generally aware of other languages, and ground their analysis of English in part on that awareness. Traditional grammarians famously took this too far, sometimes imagining English grammar as little more than a degenerate form of Latin and Greek grammar; and I'd argue that Chomksyan "theoretical" linguists take it too far in a different direction, imagining English grammar as just a special case of a general "universal grammar" built into all human brains; but even modern, non-theoretical grammars of English will generally situate various elements of English grammar in the context of other languages. (For example, the Cambridge Grammar of the English Language chapter on adjectives begins with a discussion of how to identify the word class of "adjectives" in a given language.) Since it's very common, cross-linguistically, for tense and aspect and mood to all be bound together as grammatical features of verb forms, it makes sense to tentatively adopt the same model for English, as long as that doesn't result in an analysis that's too complicated or doesn't describe the language correctly. (And even within English, some aspects of tense and aspect do get mixed up together; for example, if we shift "He says he's worked there for years" and "He says he worked there for years" to the past tense, both result in "He said he'd worked there for years." But that relates to the perfect aspect rather than the progressive aspect, and I think the main reason to regard both as belonging to a general category of "aspect" is a cross-linguistic perspective on both.)
2. The progressive aspect is, in fact, a common and basic part of English grammar. If you wanted to, you could go your whole life without needing to say something like "she began sitting"; but you can't avoid saying things like "she was sitting".
3. The most salient general characteristic of auxiliary verbs in present-day English is that they have negative forms; we have "weren't" (or "were not") and "shouldn't" (or "should not") and "don't" (or "do not"), but a proverb like "Waste not, want not" wouldn't be idiomatic in present-day English if it didn't already exist in fixed form. So although "she was sitting" and "she began sitting" both have the form S+V+PrP, we can justify regarding "was sitting" specially because it uses the auxiliary verb "was" and has the negative version "she wasn't sitting", whereas "she began sitting" uses the lexical verb "began" and has the negative version "she didn't begin sitting". Now, the fact that we regard it specially doesn't mean we have to treat it in exactly the way that we do — after all, we also say "she wasn't nice" rather than *"she didn't be nice" (cf. "she didn't seem nice") — but it at least supports the treatment of "was sitting" as a distinct construction that warrants a distinct name.

Answer 4

You are mistaken about a few things, one of which is:

since [S] - [c] - [PA] seems to be a logical breakdown of ["Jane is sitting"], and the meaning wouldn't substantially change if the sentence is interpreted that way

Let's add on the couch:

(1) Jane is sitting on the couch.

[S] - [c] - [PA] cannot be a logical breakdown of Jane is sitting in (1), because on the couch can only be correctly parsed if we treat sitting as a verb.

So before you try to figure out the concept of "aspect", I suggest you learn how to distinguish a verb from an adjective.