Contractions are pretty common in English, and I know they can be found in Spanish, French, Italian, German and Japanese (and classical Greek and Latin). But is there any language that doesn’t use them in some way? Maybe a language that uses contractions in the spoken form, but not in the written form?
I can’t think of any, but maybe someone else can?
I wonder about the many hundreds of languages that have been committed to writing only in the last two or three generations. Or even a little further back in time, like Hawaiian (that pervasive apostrophe does not mark a contraction).
EDIT: also, I’m assuming non-alphabetic written languages like Chinese are out? I wonder about syllabic systems such as many of the languages of India?
Asl?
What’s the contraction for “slap this guy in the face” in ASL? 
Interesting thought, but there isn’t anything inherently different about hand gestures that would prohibit contractions. In fact, I wouldn’t be surprised if there were more. Making the transition from one hand position to another is going to be rushed or abbreviated.
I don’t see how the writing system is relevant. Contractions are a matter of speech. That being said, Chinese languages might appear like candidates but it is not so: contractions are formed by dropping some syllables. A basic example in Mandarin: “good morning” is zǎoshànghǎo (早上好 ) but it’s often contracted to simply zǎo a(早啊 ).
What about Espreranto? Aren’t they (theoretically, at least) strict about that?
What is and is not a contraction is rather arbitrary. We call “can’t” a contraction in English because it’s written with an apostrophe and we can recognize it as being a contraction of “can” and “not.” But what happens to words after they get contracted is that eventually most speakers cease to realize that the word is a contraction anymore since it’s no longer written in a way to make its origin as a contraction noticeable and the spoken form changes enough to make its origin no longer obvious at some point.
Is ‘innit’ a contraction?
I think the the case of can’t, where the word “cannot” is also used, that we can call that a contraction without invoking an arbitrary standard. OTOH, a word like fo’c’sle, for those who even use it, is not generally known to be a contraction of “forecastle”.
What percentage of the speakers of a language would have to know that a word is a contraction before you would call it one? 100%? 90%? 50%? 10%? That’s what I mean by arbitrary. I knew that “fo’c’sle” was a contraction from the first time that I read the word. On the other hand, English speakers do not realize that “can’t” is a contraction until they are reading age, which is several years after they first learn it.
I see no reason that those of us who are literate should be guided by rules only if you have to be literate to know them. How many speakers of the English language know that the word “speak” starts with the sound represented by the letter “s”? Would you argue that the word speak does not start with the sound represented by the letter “s”, or that saying so is arbitrary?
You didn’t answer my question, which was what percentage of the speakers of a language would have to recognize that a word is a contraction before you would call it a contraction.
Whether a word begins with the /s/ phoneme is an objective matter. A trained linguist who doesn’t know the language could analyze the language and determine what its phonemes are and whether a given word began with a given phoneme. On the other hand, what you are calling contractions depends on the linguistic knowledge inside the head of each of its speakers. Whether any given speaker knows that something is a contraction depends on his internal model of his language, and that varies from speaker to speaker.
I don’t see the point of your complaint that literate speakers are the only ones who determine what a contraction is. I could equally well say that only speakers who have a complete knowledge of the etymological history of a language can determine what a contraction is. Such a speaker could say that any word that was ever created by contracting two or more words of the language is a contraction.
Thai.
What about the middle finger?
Many Asian languages, now that I think about it. Khmer, I’m pretty sure, does not. I believe Indonesial and Malay do not.
Chinese and Japanese anyone?
I didn’t initially think that Japanese had contractions, but now I have to admit that they’ll contract long names down. Like Tokyo Daigaku ( 東京大学 ) becomes Todai ( 東大 .) And one could possibly argue that having the same character ( 私 ) being pronouncable as “washi”, “watashi”, or “watakushi” as a sort of spoken contraction, even if it isn’t written.
Even more compact contractions can be found in Chinese, where two syllables are combined to form a new word altogether:
别 bié ← 不要 bú yào
这个 zhèige ← 这一个 zhè yī ge
那个 nèige ← nà yī ge 那一个
The Beijing dialect is also notorious throughout China for forming shortened words with the -r sound at the end. So, instead of the ‘standard’ 昨天 zuótiān, we get the Beijing 昨儿 zuór. This extends to a lot of expressions — you might hear 那儿 nàr for 那里 nàli, 今儿 jīnr for 今天 jīntiān, and so on.
I think the question is irrelevant. Linguists, being scientists, can determine if something is a contraction or not, based on the definition they use. Whether 50%, 80% or 100% of the speakers can tell that can’t is the contraction of cannot doesn’t alter the fact that it, objectively, is so. No less so than saying we know the etymology of a certain word even if only a tiny fraction of speakers know it, or that we know the word “speak” starts with a given phoneme.
But the collection of words which have been formed by shortening of other words/phrases is far larger than that particular fairly arbitrary group we standardly call the contractions. For example, how many people today would consider “goodbye” a contraction? I suspect most would say it is not, even though many know that it originated as a shortening of “God be with you/ye”.
I also suspect that there is not a language in the history of the Earth which has not evolved by this process of shortening of words/phrases. It’s almost as implausible as a language which has undergone no changes at all.
Ambiguities also arise in conjunction with the usual problems setting down a scheme for delineating the particular unit composing a “word”; are such common utterances as “hafta”, “wanna”, “gonna”, “gotta”, “dunno”, “didja”, “lemme”, etc., single word contractions or merely particular pronunciations of phrases? I suppose a strong argument could be made for the former, on the grounds that these words function similarly to “can’t”, where “have to” and “can not” reduce while “have two” and “can knot” do not. But once one starts going down that road, where “Jew eat yet?” opens with a “contraction”, it becomes really difficult to imagine any speaker who does not regularly, prolifically employ contractions.