16

This is a pretty basic question I guess, but anyway.

Do all (human) languages have sentences?

Most linguistic articles I read assume so, but can we take this as an assumption?

hippietrail
  • 14,687
  • 7
  • 61
  • 146
Karel Bílek
  • 269
  • 2
  • 6
  • 8
    It is a pretty basic question, but still worth asking. Of course, you have to define "sentence" before we can answer. The question to ask yourself is "What do *I* mean by sentence? – jlawler Feb 25 '13 at 05:51
  • 'sentence' is not a concept used much in linguistics. – Gaston Ümlaut Mar 30 '13 at 06:38
  • @GastonÜmlaut: Really? Not used much? Isn't the object of study for transformational syntax entirely based on the sentence? – Mitch Mar 30 '13 at 23:22
  • 1
    Not as a particular concept. The term 'sentence' is commonly used as a handy way of referring to chunks of text, but it doesn't necessarily refer to anything in particular (unless defined in some way for the purpose at hand). Other terms used in this way are: 'text' and 'utterance'. It might be that those of us who work with languages which do not have written traditions tend to prefer 'utterance'. Here's a discussion of 'sentence'. – Gaston Ümlaut Mar 31 '13 at 00:33
  • 1
    You can definitely take it as an assumption (which is not to say "yes"). In formal grammar, a language is often taken to be a set of expressions, and each of those expressions is referred to as a sentence. – Greg Lee Oct 22 '16 at 22:31

4 Answers4

7

I think even a better question would be do any languages have sentences? Sentence is an artifact of writing and punctuation. You can see how this study found it hard to compare sentence length in writing and speech for this very reason: The structure of children’s writing: moving from spoken to adult written norms. In many ways, it's not that different from asking 'Do all languages have paragraphs' where the answer is obvious.

However, all languages have some sort of a clause-type thing allowing them to express predication, attribution, etc. See Dixon's Basic Linguistic Theory: Basic Linguistic Theory Volume 1: Methodology . All languages also must have means of expressing cohesion and coherence (texture) although this is much less studied in cross linguistic perspective. Punctuated sentences are a kind of cohesive device.

Ken Graham
  • 263
  • 1
  • 4
  • 13
Dominik Lukes
  • 10,588
  • 27
  • 49
  • While I agree with you, it seems to me that linguists who study languages with a strong written tradition often talk about 'sentences', even when their examples are not from writing. Those of us who work on previously unwritten languages (I think) tend to talk about 'texts' or 'utterances'. I think this is an important issue as it relates to how some linguists have ignored the true messiness that is often found in examples of human language. – Gaston Ümlaut Jul 24 '13 at 08:18
3

All languages have sentences; both the basic building blocks (parts of speech like nouns and verbs) and the systems for constructing sentences out of these building blocks are very similar across languages.'

-Mark Aronoff (2007) Language. Scholarpedia, 2(5):3175.

So in answer to your question: YES!

Marcus Hallett
  • 31
  • 3
2

According to Merriam-Webster online, a sentence is ``A set of words that is complete in itself, typically containing a subject and predicate, conveying a statement, question, exclamation,...'' In that sense I think we could say yes.

The orthography of certain languages may not have spaces or punctuation marks but, just as all human languages have breaks between sounds, they also have breaks between thoughts. If you mean some requirement to do with predicates or subject or objects, perhaps there is a language that almost always omits one of them (I studied applied linguistics and saw many odd examples along the way). In what sense do you mean 'sentence' (as jlawler said)?

sventechie
  • 265
  • 1
  • 8
0

If we take a sequence of different parts of speech (or a sequence of different words) as a basic definition for a sentence, then there are two types of languages which, presumably, have no sentences falling under the definition:

a) polysynthetic and/or incorporating languages (like Chukchi, Bella Coola or Tiwi) for their word-sentences, and

b) languages with vocabulary where words cannot be divided into parts of speech, like Zhuang or Chinese.

Gaston Ümlaut
  • 6,697
  • 30
  • 46
Manjusri
  • 2,781
  • 1
  • 19
  • 27
  • Whoever downvoted, be sure to leave a comment to help the answerer learn what makes a better answer. – Joe Mar 28 '13 at 20:14
  • I think this is a trace from some unskillful person I know from another site. It's just a smut which does not have anything to do with the quality of my answer. A kind of downvoter who has nothing to say due to the lack of any knowledge about the subject. – Manjusri Mar 29 '13 at 06:06
  • I didn't downvote but I'm tempted. Re a) polysynthetic languages still have sequences of morphemes. The fact that the morphemes don't have space between them doesn't mean there aren't sentences. b) These languages do have part of speech categories eg see this paper. – Gaston Ümlaut Mar 30 '13 at 02:05
  • a) The morphemes are parts of words, not of sentences (pls see the initial definition above). b) The parts of speech in isolating languages (including Zhuang) are defined by what can be best named 'probabilistic semantics' (pls see the Bodomo's paper on pronominal system in Zhuang, or a paper on 'can' representation in the same language). – Manjusri Mar 30 '13 at 03:42
  • I think it's important to be aware that 'sentence' is not a concept used much in linguistics as it pertains best to writing. I'm not sure what you mean by saying the POS in isolating languages are defined using 'probabalistic semantics', POS are defined in pretty much the same ways across all languages and Zhuang (according to Bodomo) has a pretty unsurprising set of POS categories. – Gaston Ümlaut Mar 30 '13 at 06:32
  • POS are defined conceptually and they are not same in languages of different types. For probabilistic semantics, pls see this link. http://www.illc.uva.nl/LaCo/CLS/presentations/lappin-vaneijck11cls.pdf – Manjusri Mar 30 '13 at 07:17
  • And here is another paper on 'POS' in Zhuang http://www.google.ru/books?hl=en&lr=&id=n9p4rl09ec0C&oi=fnd&pg=PA221&dq=morphology+of+zhuang+language&ots=_IsPxsEUNf&sig=eqho04NN-XMFkQUXLtzq4av55zE&redir_esc=y#v=onepage&q=morphology%20of%20zhuang%20language&f=false – Manjusri Mar 30 '13 at 07:19
  • Perhaps you misunderstand me; I'm not saying the same POS are found in all languages, I'm saying that in linguistics the same broad methodology is used to determine POS regardless of the language: by examining distributional and combinatorial properties (ie morphology and syntax). You say they are 'defined conceptually'; if by that you mean to say they are defined by semantics, that is incorrect. – Gaston Ümlaut Mar 30 '13 at 13:47
  • Thanks for the paper on the ACQ element in Zhuang. But it does make clear that in Zhuang words can be divided into POS categories. – Gaston Ümlaut Mar 30 '13 at 14:03
  • In my opinion, it makes clear that 'the behaviour of ndaej in Zhuang in the property of the element itself'. I have an interesting book on Chinese grammar (another isolating language) by Kurdyumov further explaining the arbitrary and probabilistic nature on so called 'parts of speech' in isolating languages, but that book is in Russian. – Manjusri Mar 30 '13 at 15:13
  • 2
    This is getting too long a thread so this is my last comment, but: yes /ndaej/ is an oddity in Zhuang and may be best analysed as its own POS. And I think it's fair to say POS are arbitrary and probabilistic in all languages, but my point all along has been that Chinese and Zhuang do have parts of speech. – Gaston Ümlaut Mar 31 '13 at 00:18