dialect, noun = a form of a language which is peculiar to a specific region or social group (from: Compact Oxford English Dictionary). Another definition, from Wikipedia, is “a complete system of verbal communication (oral or signed but not necessarily written) with its own vocabulary and/or grammar.”
Origin: from the Greek διάλεκτος (dialektos) – discourse, way of speaking.
Deciding whether a particular form of speech is a language or a dialect is a task fraught with difficulties. As well as linguistic criteria, there are also political, geographic and cultural issues to be considered. For example, closely related languages spoken in different countries, such as Norwegian, Swedish and Danish, might be considered dialects of one language if they were all spoken in a single country. There is considerable mutual intelligibility between these languages, but each of them has its own written standard, or two written standards in the case of Norwegian, which seems to be a good criterion for distinguishing languages. Perhaps you could define a language as a dialect with a standardised written form.
In the case of Arabic, there is one standard written form: Modern Standard Arabic, and many colloquial spoken forms, all of which are considered dialects of Arabic and are rarely used in writing. Though there is a high degree of mutual intelligibility between neighbouring colloquial forms of Arabic, for example, Jordanian and Syrian Arabic, speakers of these dialects have difficultly understanding the Arabic spoken by Moroccans and Algerians. This is an example of a dialect continuum.
Where do dialects come from?
When groups of people are isolated from others, the way they speak tends to drift away from mainstream forms of their language. Changes in the mainstream forms may not occur in the isolated form, and vice versa. Over time, the isolated form develops into a distinctive dialect, and if the isolation continues for long enough, that dialect may eventually become a language with it’s own written standard.
Another source of the differences between dialects comes from the languages once spoken in the regions where the dialects are now spoken. When people shift from one language to another, they usually carry over some features of their original language, including pronunciation, vocabulary and grammar. The dialects of English spoken in Ireland, for example, retain quite a few features of Irish. Similarly, substrates of the languages once spoken in the different regions of England can be found in some of today’s regional dialects. Moreover, the modern Celtic languages may contain substrates of the languages they replaced, whatever they were.
Another great, objective article 🙂
I heard once, that the mutation in the beginning of the words in Celtic languages is in fact a feature that was originally in the language of pictish people! Don’t know if this is true or not, but by my humble experience with Breton … I think there are no initial mutations for the words … are there any?
By the way, maybe a good example of a dialect that turn into a language …. I think Manx and Irish Gaelic (with Scottish Gaelic) can do the job!
What constitutes a “dialect” really depends on the language, too. Many of the “dialects” of Chinese would actually be considered separate languages, were they not written identically with ideographs. Similarly, the degree of diglossia in Arabic is pretty (though not quite that) dramatic. A lot of German dialects can be mutually unintelligible, at least in part. My German teacher had a friend who was a physician who ran into a monk in the Black Forest. The monk (as well as all the other monks in his monastery) were from the local region and only spoke the local dialect among each other. They understood Standard German (Hochdeutsch), but didn’t speak it. And the doctor didn’t speak the local dialect. So, there in the late twentieth century, these two educated Germans had a conversation in their common language. Latin.
TJ – there are four types of initial mutations in Breton: soft, spirant, hard and mixed. For example, p changes to b or f, and t changes to d or z.
Manx and Scottish Gaelic were originally dialects of Irish and there is still considerable mutual intelligibility between them, particularly between the Irish spoken in the north of Ireland (Ulster/Donegal) and the Scottish Gaelic spoken in the southern parts of the Hebridies.
For example, ‘How are you?’ in the Ulster dialect of Irish is “Cad é mar atá tú?”*, and it’s “Dè ma a tha thu?”** in southern flavours of Scottish Gaelic.
* In Munster it’s “Conas atá tú?” and it’s “Cén chaoi a bhfuil tú?” in Connemara.
** “Ciamar a tha thu?” is the ‘standard’ version
The use of cha as a negatator is common to Ulster Irish, Manx and Scottish Gaelic – e.g. I don’t understand = Ní thuigim (Standard Irish), Cha dtuigim (Ulster), Cha nel mee toiggal (Manx), Chan eil mi ‘tuigsinn (Scottish Gaelic).
Thanks for the info!
By the way, I have here an Irish teaching book and it’s from “teach yourself” series, and I don’t know if they actually use the standard version (one after reform in 1945 I think) but they use “conas atá tú” for “how are you.”
That was quite load of information about Breton. The problem with Breton is that I really can’t get any english gate to learn or at least view some information about this language! Most of the resources are in french! And believe me …. if it wasn’t for Denez Prigent, I wouldn’t ever have heard of it!
Many of the “dialects” of Chinese would actually be considered separate languages, were they not written identically with ideographs.
I’d like to bring up a few points. First, Chinese characters don’t represent an ideographic form of writing. Second, the various Sinitic languages are not written identically; rather, speakers of these various languages are taught to write Mandarin, which is a very different reality than the universality myth of Chinese characters.
When written in Chinese characters, Taiwanese and Cantonese (two languages for which Hanzi scripts exist) are quite different than written Mandarin. Speakers of one language can make out some of a text written in another language; but that doesn’t have anything in particular to do with Chinese characters. (Consider, for example, how someone in Italy can read parts of a newspaper printed in Madrid even if that person doesn’t know Spanish.)
Mark S. – I followed your link about the objection to the term “ideograph.” It was educational and disabused me of the illusion that Chinese words are “characters.”
The writer asserted that ideographic representation of all ideas would be impossible for humans to master because of the sheer number of symbols that would be needed. That seems reasonble enough. But, my question for anyone who knows is: wouldn’t Sign Language, though not written, represent a system of communicating ideas apart from any relation to sound? They don’t seem to be limited in communication.
How does sign language convey all the ideas we take for granted in sounding out? Do signers “spell” with their hands a lot?
In reply to Mark S’ post, being but an amateur learner of Japanese, I do not dare to contradict John DeFrancis’ in-depth analysis (I might be stricken by lightning if I did so), but I cannot refrain from raising a doubt about its validity concerning the language I am studying. In particular, my attention was caught by the passage in which the question “When is a pictograph not a pictograph?” receives the answer “When it represents a sound”.
As everybody knows, Japanese makes use of Chinese glyphs, or kanji. In this language most kanji do suggest a concept (not simply a sound, nor a word), and sometimes even more than one. At the same time, kanji with completely different shapes sometimes convey very similar meanings, whose slight difference is used for rendering nuances. An example will make this clear.
The verbs “to write” and “to paint” are expressed in Japanese by means of the same verb, “kaku”; since a fude brush is traditionally used for both activities, this analogy also semantically suggests that kanji are indeed painted images, yet strongly stylized. The first of the two meanings (to write) is rendered by means of kanji 書, whose top part (radical 129) is the stylized shape of a brush, and its bottom part (radical 72) means “day” or “sun”, as writing is an activity that requires light. The other meaning (painting) is spelt with kanji 描 , whose left half is the stylized form of radical 64 (“hand, arm”) and whose left part points to flowers (radical 140) and rice-paddies (radical 102), i.e. a typical subject for landscape paintings. However, taken individually, the two kanji (書 , 描) have no specific meaning nor reading, and in fact cannot stand alone in a sentence, but only followed by either okurigana (書く, 描く = ka-ku), or by another kanji so to form a compound or jukugo, in which case they often no longer sound ka-, but sho and byō, respectively, and their shade of meaning may further change. For instance, 書店 shoten (individually: “writing” + “shop, store”) actually means “bookshop”, and 描写 byōsha (individually: “drawing, painting” + “copy, photograph”) means “description, account”.
Therefore, a kanji does not describe a sound, nor a single word, but a number of words bound together by an association of ideas.
In Japanese, very few kanji are used for a purely phonetic purpose, and their reading is always on’yomi, or Chinese-derived.
Speaking of the Chinese system of writing, John DeFrancis’ book says: “its characters do not represent ideas, but words”, and quotes Boodberg: “Signs used in writing, however ambiguous, stylized, or symbolic, represent words”. Mark adds: “Chinese characters don’t represent an ideographic form of writing”.
The aforesaid examples show that these statements cannot be held valid for Japanese.
Hi, Polly:
One thing to keep in mind is that there is no universal sign language, which might be expected if full “ideographic” communication systems were possible. Sign language is an interesting question. I wish I knew more about it.
++++++
Ciao, Andrea:
In terms of your specific examples:
The first of the two meanings (to write) is rendered by means of kanji 書, whose top part (radical 129) is the stylized shape of a brush, and its bottom part (radical 72) means “day” or “sun”, as writing is an activity that requires light. The other meaning (painting) is spelt with kanji 描 , whose left half is the stylized form of radical 64 (”hand, arm”) and whose left part points to flowers (radical 140) and rice-paddies (radical 102), i.e. a typical subject for landscape paintings.
These are not real etymologies but are instead good examples of the fairy tales that pass for scholarship in so many works. The “sun” that you point to was originally 者 (zhe in Mandarin), which was the phonetic element of the character; although the form has changed over the years, that doesn’t change the fact that “sun” has nothing to do with this but a phonetic element does. Similarly, in 描, the right side of the character (苗) does not represent the idea of flowers and rice paddies but is instead a phonetic element (miáo in Mandarin).
The fact that Japan borrowed these and other characters at various points in time and that pronunciations have shifted or others have been applied does not alter the fact that these are not ideographs. And that Japan uses a horrifically complicated orthography doesn’t really change anything either. Having more than one possible pronunciation is not at all the same thing as there being no associated sound or meaning being transferred supra-linguistically. (Consider, too, that English has spellings that are pronounced different ways and mean different things, depending upon context, e.g., “read,” “protest.”)
Since you point to Japanese, I’d like to refer you to the works of Prof. J. Marshall Unger. Confusing Language with Writing and The Modern Japanese Writing System would be good places to start.
And always try to keep in mind that language is not a representation of writing. Writing is used to represent language. Language first, writing later. All by itself that’s largely enough to explode the ideographic myth, as Peter DuPonceau noted nearly 200 years ago.
I express my appreciation to Mark, whose scholarly explanation I do acknowledge, but I’m afraid that my doubt is still lingering. I’ll make another very simple example.
The word “adult” can be translated at least in two ways, with the usual choice of nuances: 成人seijin (literally: “turned into a person”) suggests the idea of adult as an over-18, i.e. in opposition to an underage, whereas 大人 otona (literally: “big person”) suggests the more generic idea of an adult vs. a child. The second word is the graphic representation of the very concept of an adult, without the least relation to the standard sounds that either 大 (tai, dai, oo-) or 人 (jin, nin, hito) may take in any other word. If the theories of John DeFrancis’ and others were true, there is a wealth of other kanji bearing the standard sounds “o + to + na”, or “oto + na”, etc. which could have been chosen to spell this word, yet the ‘visual’ concept would have been lost. Obviously, “otona” this is not the only word that completely differs in pronounciation from the individual kanji it contains, but points to the visual representation of the meaning it conveys, either actually or metaphorically.
Whether these words too may be explained with the archaic shape and/or sound of the original Chinese glyphs (if I can dare to still call them ‘glyphs’), I honestly don’t know. But whenever I ask similar questions to Japanese friends, however deep their knowledge of their own native language may be, they too fail to give an answer. To me, this means that even accepting John DeFrancis’ theory, to Japanese native speakers (and to foreign learners like me) the language de facto contains ideographs.
Maybe we could agree on the fact that Japanese is a *partially* ideographic language, whose written form, in my own experience, is rather loosely bound to its spoken form: some jukugo can be read in two completely different ways, sometimes with the same meaning, sometimes not.
Maybe we could agree on the fact that Japanese is a *partially* ideographic language….
Sorry, I can’t agree with that at all.
That your Japanese friends don’t know the true linguistic situation behind this is not at all unusual. Native speakers of a language — any language, not just Japanese — are not endowed with an innate scientific understanding of the script(s) used to write it. And just because some of them might be under the misconception that kanji are ideographic doesn’t make it so. (Interestingly, Japanese and Chinese seem to have picked up the ideographic myth from Westerners, esp. Ernest Fenollosa.)
To address your example, first, it is crucial to keep in mind an elementary but often overlooked distinction: the kanji “大人” are not the Japanese word for “adult” but a way of representing in writing the Japanese word for “adult.” As I wrote above, “language is not a representation of writing. Writing is used to represent language.” Kanji is not a language; no one speaks kanji. Consider the English word “petroglyph.” Petro does not need to be written with a Chinese character used to indicate “rock” and glyph does not need to be written with another character used to indicate “writing” for the word petroglyph to have meaning. The word means what it means regardless of the script used to write it; this is no less true of Japanese words than those in any other language.
The varieties of approaches to pronouncing some kanji are the result of differences in the time, method, source language, etc., of the adoption/application of kanji to words in Japan. Languages change over time. Pronunciation changes over time. Meanings of words change over time. Add to that the fact that sometimes Sinitic pronunciations were borrowed along with the characters, sometimes not. It’s a real mess.
That just means kanji are difficult and not a particularly good fit for modern Japanese, which is not even in the slightest way the same thing as calling them “ideographic.”
Maybe we should start again the debate from the question ‘what is an ideograph, and what does ideographic script mean’.
John DeFrancis’ text says: “Aren’t Chinese characters a sophisticated system of symbols that similarly convey meaning without regard to sound? Aren’t they an ideographic system of writing? The answer to these questions is no.”
Considering the spelling of 大人otona, I absolutely agree with you that this is a written representation of the word. This means that 大人, which according to the standard phonetics of Japanese kanji should be pronounced either “ōjin”, or “daibito” – how many times I made this mistake before “otona” became automatic! – conveys meaning without regard to sound, in open contradiction to what DeFrancis maintains.
To me, an ideograph is a stylized graphic representations of a given meaning, either expressed by individual kanji such as 休 (a person by a tree = “rest, day off from work”), or by compounds such as the aforesaid 大人. Once a graphic shape conveys a meaning, sound loses importance, because in different languages the word it points to is not pronounced in the same way. In this, kanji and jukugo are not much different from pictograms used in international signs, such as an arrow or a finger pointing towards a given direction, or an exclamation mark to signal an alert, etc. In fact, as you observe in your first post of this thread, without knowing their Japanese reading, a Chinese would understand their meaning in eight or nine cases out of ten (yet still running the risk of misunderstanding them, see my other post in the thread at
http://www.omniglot.com/blog/2006/04/08/jidouhanbaiki/ ).
Furthermore, Japanese has a dual spelling system, i.e. kanji and kana (phonetic syllables): when kanji are used rather than kana, the primary clue conveyed by the spelling is the meaning of the word, not its sound, whereas a spelling with kana has the opposite effect.
I once again underline that my considerations refer exclusively to Japanese, not to Chinese.
I’m not well up on linguistics, but it seems to me that more pernicious than the ‘ideographic myth’ is the ‘phonetic myth’ that’s grown up to replace it, apparently the work of John DeFrancis, J. Marshall Unger and the racist William C. Hannas. There is no reason that a writing system need be commutatively equivalent to a speech system, any more than sign language need have anything to do with speech. All three, I’d say, are potentially independent but rely on the ability of the human brain to manipulate syntagmatic chains composed of discrete units of meaning.