In english, a 'complete sentence' seems to refer to having at least a single, complete clause — i.e. a subject (noun) and verb — e.g. "I run". This seems to be engrained in the concept of a complete idea or thought, basically a classical predicate — providing information about a subject.

The concept of a "complete" sentence (or complete idea) is definitely common in other languages (at least latin and germanic). Is it common to all languages and cultures?

At first it may seem like an arbitrary linguistic construct that need not be obeyed in unrelated languages, but at the same time --- as I tried to suggest above, it seems to be deeply rooted in our way of thinking.

Edit: I hoped it was obvious from the context I gave above, but I'm not interested in the pedantic question of what minimal requirements satisfy the purely-structural concept of a 'sentence'. E.g. perhaps, "I.", can be a complete sentence in response to a question, or "Go.", as an imperitive. This related, but different, topic is the basis for all of the discourse I've found online, and what I've seen discussed at length by scholars (e.g. Chomsky). Instead, I'm interested in the the concept of a 'complete sentence' in the context of 'a complete thought' per se, in the context of theory of mind/thought, compared in different cultes and languages.

  • 2
    Well, not really. – ScotM Apr 2 '15 at 15:20
  • No. (That was a complete thought, expressed in a complete sentence, using one word.) – Dan Bron Apr 2 '15 at 15:39
  • Are you asking whether the notion that a sentence can be incomplete (i.e., not contain enough information to represent a ‘complete idea’) is universal? If so, I’d say yes, it is. What different languages count as enough information to represent a ‘complete idea’ varies, of course; but the notion that an idea can be either ‘complete’ or ‘incomplete’ is surely a universal artefact of the practical needs of communication. – Janus Bahs Jacquet Apr 2 '15 at 16:25
  • Seems like you're asking for the linguistic/syntactic analog for Well-Formed Formula (WFF), which is a logical concept. A well-formed predicate calculus formula is a Predicate with 0, 1, 2, or 3 Arguments, depending on the requirements of the Predicate itself. See the first page of the Logic Study Guide. This is the semantic/logical concept that's usually linked in "grammar" school with "complete sentence". – john lawler in exile Apr 4 '15 at 3:18

Some languages allow a sentence to consist of a predicate only, without any explicit subject -- not just in contexts like a response to a question, but completely independently. E.g. Hebrew:

Kar. 'It's cold.'

This is literally just 'Cold'; the dummy it of English has no equivalent in the Hebrew sentence.

| improve this answer | |
  • Thanks! This is a really interesting example. Is the verb implied here? I think that's important. – DilithiumMatrix Apr 3 '15 at 13:41

Both the concept of 'sentence' and 'complete idea' are a result of the pedagogical grammar tradition.

The fact that simple Subject-Predicate clauses are given as examples in the European linguistic tradition. If they chose, imperative sentences like 'Hand me the keys!', the question would not arise.

The notion of 'complete thought' is really just heuristic for identifying the minimal semantic core of a sentence. E.g. "A man walks." and "It rains." vs. "A wise man walks very quickly when it rains." But it is clear that the constituent ideas of the above sentence are not complete with respect to the speaker's intention, nor are they complete when compared to the subjectless command "Sit!".

In fact, you could argue that "Sit!" expresses a more complete idea than something like "a man walks" which rarely occur on their own.

So, as you can see, the question does not even make sense in English, let alone in a comparative perspective.

However, setting the complete idea thing aside, it seems that the subject-predicate structure of a basic declarative clause is present in all languages although the grammar of this can be quite different from English (as in ergative languages). See Dixon's 'Basic Linguistic Theory' for more details. But it would be dangerous to try to make much more of it than the fact that by definition declarative clauses are there to say something (predicate) about something (subject). [Edited to add] This does not mean that the simple declarative clause is in some way basic to language or the only way to express ideas in language. But in many contexts (logic, philosophy of language, school grammar) people are used to talking about language as if expressing simple predicates was what the rest of the system was growing from.

| improve this answer | |

I suppose all human languages must have some way of conveying complete ideas, if that means saying something which, if translated into English, would have a subject and a predicate. But really, that seems to be more a fact about English than about other languages, since it's about how ideas are expressed in English. If you look specifically at the overt forms of sentences, there doesn't seem to be any real evidence that the subject-predicate form is universal among human languages.

For one thing, there are plenty of sentences expressing a complete thought in English and in other languages that do not have the subject-predicate form. And for another, there are many languages which have VSO word order -- that is, a basic sentence form in which the subject comes between the verb and the object, so that there can't be a single predicate phrase like the verb phrase of English, which contains both the verb and the object.

The Wikipedia has a short entry on VSO languages: Verb subject object. From that article:

Examples of languages with VSO word order include Semitic languages (including Arabic, Classical Hebrew, and Ge'ez (Classical Ethiopic)), and Celtic languages (including Irish, Scottish Gaelic, Manx, Welsh, Cornish and Breton), and many Mesoamerican languages.

| improve this answer | |
  • Thanks for your response. I don't think the ordering (VSO vs. SVO) changes the concept that you still require a subject and verb. Additionally, it was my understanding that all romance languages had the same 'complete sentence' requirements; I think you could make an argument that the very nature of a conjugated verb makes it a subject-verb combination, and thus even more requisite! – DilithiumMatrix Apr 2 '15 at 17:51
  • I don't understand your reasoning about conjugated verbs at all. Conjugated verbs in many romance languages (among many others) have affixes on the verb that make it easy to dispense with the separate subject phrase that would be required in English. – Greg Lee Apr 2 '15 at 18:41
  • 3
    I don't see how word order has anything remotely to do with this question. – fdb Apr 2 '15 at 18:53
  • @GregLee exactly: you don't need to add a subject because it's effectively included in the verb itself. In spanish, if you say "voy" it doesn't just mean the english "go" --- it specifically, and explicitly, means "I go". – DilithiumMatrix Apr 2 '15 at 19:33

A lot of languages have sentences without a verb. In Latin you can omit the verb “est” (is) or “sunt” (are) and simply juxtapose the subject and predicate. Some languages nearly always omit “is” in the present indicative: Russian is one and Arabic is another.

| improve this answer | |
  • Is it not implied in these examples? – DilithiumMatrix Apr 2 '15 at 19:31
  • I would say no. If there is no word for "is" you cannot imply it. – fdb Apr 2 '15 at 20:20
  • Who said Russian and Arabic don't have a word for "is"? – Nikolay Ershov Apr 2 '15 at 20:37
  • есть means “there is” and is not normally used a copula. yakūnu is used to refer to future action (“will be”) and does not mean “is”. The subjunctive yakūna and jussive yakun are used to mean “is” in various subordinate clauses, but that is why I wrote “in the present indicative”. (Sorry to insist on this, but I do actually know Arabic very well.) – fdb Apr 2 '15 at 20:53

For starters, you need to distinguish "complete" things from "incomplete" things. Supposing that my intention is to say "Well, I hope that the next time you go to that store, you'll count your change", but after saying just "well, I", I'm interrupted. That might be an "incomplete sentence". It's not actually a sentence, it's part of a sentence, so you have to understand "incomplete sentence" as being a fragment of an intended sentence. The intended sentence isn't necessarily a complete thought, since I was also thinking that this store cheats people in giving change, and maybe some day the law will get after them, maybe even closing the store or prosecuting the owner, but I didn't intend to say all of what I was thinking.

The concept of "a thought" is so hopelessly nebulous that it's really pointless to talk about "thoughts" and their relationship to linguistic form, and instead we focus on something that we understand a bit better, namely propositions. In that case, we might ask questions like "do all sentences convey propositions?", or "do all linguistic utterances convey propositions". Many utterances don't count as sentences, by standard syntactic accounts of what a "sentence" is. For example, if I ask "Who took the sandwich" and you respond "Bill", that isn't a sentence. It can be contextually interpreted as standing for "Bill took the sandwich", or "I think that Bill took the sandwich", or "Bill is the one who took the sandwich", or any number of other actual sentences.

As far as I know, there is no language where speakers are incapable of generating well-formed and intended utterances which are less than a sentence: bare NPs are always possible utterances, and are often used. People don't usually walk into a room and say, with no prior context, "The old grey mare", but if we automatically convert all utterances into "intended sentences" just in case they don't actually qualify as actual, full, well-formed sentences, then we are just begging the question of the relationship between sentences and propositions -- we are making sentences out of things that aren't sentences, because we can guess what the underlying proposition is. If we don't interpolate "missing parts" (syntactic deletions and other omissions), then in ordinary speech, most utterances are probably not complete sentences. Yet most often, the intended proposition is easily reconstructable. Therefore, I conclude that complete sentences are not a significant desideratum in communication.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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