How Linguistics Can Help You, Part 1: Articulatory Phonetics
Linguistics is a topic that is often misunderstood, either through oversimplification (well, it’s just languages, right?) or overcomplication (it’s full of symbols and too hard for me). This series of posts is intended to make Linguistics user-friendly for writers, by relating linguistics topics to the use of created languages in science fiction and fantasy. If you are digging into creating an entire language that penetrates an SF/F world, this should certainly apply! But they should also be helpful for those who are using minimal-penetration languages in SF/F – languages that consist mostly of names for people and places.
My first topic here will be articulatory phonetics, which deals with how the human vocal tract creates sounds.
Knowing the principles of how the vocal tract works can help you create languages that follow naturalistic patterns of pronunciation, thus making created languages that seem more natural.
One of the key assumptions in the discussion that follows will be that we’re working with a species which, like humans, can perceive vibrations in the air (whether through ears and hearing or by other means like antennae). While this does restrict us somewhat, it still allows for a lot of possibilities.
Let me begin with a caveat before we begin our tour of the vocal tract. If you’ve never studied linguistics, this may appear complex – but it’s not as bad as it seems. Just because there are a lot of variables you can change about a language doesn’t mean you should go about trying to change them all.
Okay, so here we go:
1. Powered by the diaphragm, the lungs emit an airstream that can be shaped by other parts of the vocal tract. This is the power source for the sounds. Change this element, and you’ll have a drastically different language, but one that will be a bear to transcribe into English!
2. The vocal cords can vibrate when the airstream passes by them. All vowels are “voiced” sounds, i.e. sounds with where the vocal cords vibrate. So are consonants like b, d, z, v, y, l, r, n, and m. In the case of an alien, it’s important to know that this creature possesses something like vocal cords, or at least something able to create a consistent humming vibration, if you’re going to use any voiced consonants in transcribing its language. Language sounds without this vibration are called “unvoiced.” Whispering is entirely unvoiced.
3. The mouth is a resonating space for vibrating air. In human languages, the quality of vowels is altered when the tongue is used to alter the shape of the mouth space. The position of the tongue is described in two dimensions: height (high, mid, low) and front/back. Here are some examples of the position of English vowels.
[i] as in “feet”=high front [u] as in “hoot”= high back
[E] as in “bet”= mid front [o] as in “boat”= mid back
[ae] as in “hat”=low front
If you go into any beginning linguistics textbook, it’s easy to find a graph of the mouth space and the vowels involved; you can also Google “vocal tract.” Here, I’d prefer to talk about what to do with them. If you have an alien, try to think about the kind of resonating space it uses to create speech sounds – the length of its muzzle or other factors might change things significantly. You can also think about how it might change the shape of that resonating space (with tongue or other muscles), because this would affect its ability to pronounce human languages.
If you have a human or fantasy human, the problem is easier, but you can still think about how the language pattern might use vowels with different characteristics. Do your people generally avoid mid vowels? Avoid back vowels? Do they tend to pronounce vowels across a word with the same kind of mouth and tongue position (say, making all vowels in a single word high)? Do they generally keep their lips rounded or unrounded? There are lots of options here.
4. The air flow can be stopped or blocked in different ways by the tongue, teeth, and lips. When the air flow is blocked completely, that’s called a stop (for example, p/t/k/b/d/g). When it’s still flowing but partially obstructed, that can be called a liquid (l/r), an affricate (ch/ts) or a fricative (s/th/f/z/v). W and y are called glides. Consider the “tools” your creature or person has for altering air flow. Where will most of the obstructions occur? Far back in the mouth near the uvula, as with French R? In the front with the lips? At the alveolar ridge behind the teeth where we create sounds like t/d/s/z?
There’s more I could talk about, but I don’t want to go overboard…
In fact, you’d be surprised how few things you need to change to give an entirely different flavor to the alien words you use. Here is an example of a language that I created for my story, “Cold Words” (Analog, October 2009).
I had an alien with a long muzzle and tongue, so I decided that there were a lot of different kinds of “l” and “r” sounds in this language. In English I decided to use single “l” versus double “ll” and single “r” versus double “rr” to indicate these sounds, even though I didn’t know exactly what they sounded like. I also decided to avoid all unvoiced consonants – mostly for the sake of argument, and for giving the language a distinctive “feel.” That means plenty of m/n/d/g/b/v, etc., but no p/t/k/s/f.
To my mind, the biggest advantage of using principles of articulatory phonetics is this: if you use natural language patterns to guide your choices, the resulting created languages will seem less arbitrary and more convincing.
How Linguistics Can Help You has appeared previously at TalktoYoUniverse and at the blog of the SFWA.
This is such a great topic! However, I feel that trying to use this information to make a conlang seem “less arbitrary and more convincing”, natural but distinct, is always going to fall short simply because the thousands of years of articulation/perception balance isn’t there when you’re creating a language. But what you mention about changing the morphology of an alien or a fantasy human (e.g. longer nose, no nose, larger jaw, etc.) is where this articulatory knowledge can really be helpful. With a non-human body, there is the opportunity for non-human sounds and subsequently the interaction of beings with disparate production/perception systems. The fun begins! I look forward to the rest of this series!
Leon, I see your point about natural language evolution, but it is possible to look around the world and take examples of phonological trends from languages there, thus borrowing a bit of naturalistic language evolution. Fortunately, unless you’re designing a language like Na’vi, you don’t usually have to have lots of it in use in your story. The general sense that a language “fits with itself” is generally enough to get readers through a story. I’ve always enjoyed playing with the phonetic effects of alternate physiology – thanks for your comment!
Great post, Juliette. I’m not a linguist, as you know, but even so it annoys me when sf/f writers just pour a jumble of letters onto a page (usually with a heavy sprinkling of apostrophes) without the slightest attempt to make them sound coherent.
One trick I use, as an amateur, is I collect dictionaries, especially non-European, and then for a story draw several words from the same language. Turkish and African languages work well I find. Often I’ll also draw from a couple of different languages, just to add a little realistic salt…
I had reason to think a lot about this when writing The Biology of Star Trek and inevitably had to discuss not only the Universal Translator but also that amusing pidgin, tlhIngan Lingan-Hol… Klingon — which is far less “alien” in sound and syntax than its monolingual creators (like to) think.
Unless the author is very dedicated, talented, or both, constructed languages usually not only lack internal coherence and depth, but are also “pristine”, without any of the arbitrary flourishes that commonly exist in real languages that evolved and interacted with others.
Also, no matter how carefully an author constructs a language, there is another issue involved — how English speakers pronounce vowels, and where they place the default accent (always on the penultimate syllable, which makes a hash of my name, to give one example).
Almost all other European languages are devoid of pitch for either sound or meaning and pronounce vowels “whole” — a single value per letter, fully voiced. Not so in English. As a result, even if someone puts a lot of attention in names and terms, it is not certain that they will be voiced as intended by the readers. At the same time, because of its almost complete lack of inflected endings, English is very good for coining neologisms.
A bit more on the topic of language in speculative fiction:
Jade Masks, Lead Balloons and Tin Ears
http://www.starshipreckless.com/blog/?p=1959
Oops! tlhIngan Hol is the correct term — somehow I managed to duplicate the first half. Cause for a blood feud in that culture, no doubt! *laughs*
I don’t believe the creator of Klingon, Marc Okrand, is monolingual – he’s a linguist who’s worked on West Coast Native American languages. Klingon enthusiasts, on the other hand, seem to be disproportionately monolingual English speakers who carry a lot of unquestioned assumptions into the framework Okrand created.
This is an interesting observation – I’m familiar enough with Marc Okrand to know that he and I both spent some time at the UCSC linguistics department, though I don’t know much about his history with speaking foreign languages. I think you make a good point, Tracy, about people bringing assumptions from their native language into their use of a second language (in this case, Klingon).
Athena, I agree that it’s very difficult to create a truly foreign language. What you call “pristine” is what I call “well-behaved” – it’s how we often like our languages to work so we can keep track of them, I think. I have always had the sense that Tolkien’s elvish language was quite well behaved, at least inasmuch as the character names had quite specific meanings in the language (which is a perfectly legitimate and interesting choice). I’m glad I haven’t yet been put in a position where I have to create a usable language with a full vocabulary. I suppose I might try to create some of the little chaotic language whorls that don’t conform to the main body of the “rules.” It’s certainly something I’d enjoy seeing in a conlang. However, as far as having a speaker (of any language!) pronounce such languages “properly,” I don’t really hold out hope. There’s far too much that can’t be expressed in orthographic form. I just figure I know how it’s pronounced in my head, and that others will pronounce it as they like. This ends up generally being in accordance with the rules of phonology for their own language – and that’s fine with me. I find if I try to anticipate mispronunciations, I stress out too much!
There’s no question that the issue of pronunciation is particularly fraught for tonal languages — and English qualifies as tonal, the sole European language in that category. Add dialects and regional accents to the mix and it becomes quite the word salad!
English is not tonal. Swedish is a lot more tonal than English is — at least, it has a tone accent, which English does not have. Maybe you’re confusing phonetic tone and intonation, but even there, even though English intonation is quite pronounced, other European languages have similar tone contours.
I am not a professional linguist, so I lack the specific vocabulary that would lend precision to this exchange. Also, I should have qualified my comment to say “of the 4.5 European languages I know” — which do not include Scandinavian or the south Serbian family, said to be tonal. English is tonal insofar as cup and cap sound essentially identical, yet mean something very different (to a Greek who starts learning English, both are “kap” and it takes a while to distinguish them).
“Cup” and “cap” have different vowels; there is no “tonal” difference between them. I’d suggest you take a look at something like this (http://en.wikipedia.org/wiki/Tone_(linguistics)) to clarify what you are thinking.
I know what I’m thinking, thank you.
If you’re thinking that ‘cup’ and ‘cap’ are different words, then I think everyone is on board. If you think that they are different words because of tone, you are wrong. Swedish=tonal. Serbian=tonal. Latvian=tonal. English=not tonal.
And speaking of clarification, my note didn’t say that the two words have the same vowel, but that they sound essentially identical to a non-English speaker. As identical as two Chinese words of a different pitch sound to a non-Mandarin/Cantonese, etc. speaker, until their ear learns to discern the difference, which carries meaning with it.
Yes, that’s all fine and good, but the issue is not one of pitch or tone; the words have different vowels.
As Christopher Doty has said, that difference is not tone. English distinguishes between “cup” and “cap” based on vowel quality – the tongue is in a different position for the two. Informally, they’re different vowels.
It’s true that it’s difficult for speakers of languages that don’t have that particular distinction to hear the difference, much as English speakers have difficulty hearing the difference between aspirated and unaspirated consonants in Hindi. But that doesn’t mean the difference between two different vowels is an aspiration difference, any more than it means what you certainly seem to be claiming – that it’s a tone difference. Tone is a distinction between the pitch patterns of the same vowel, and “cup” and “cap” are not distinguished by different pitch patterns.
Chris, you keep insisting that the two words have different morphemes (which is obvious), while I have repeatedly explained that my point is that the phonemes they use sound identical to an outsider until her/his ear learns to distinguish the difference.
This point, incidentally, does not come from me alone. I have heard it from others whose first language is not English and who are used to “absolute” vowel values (Greek, Spanish, Russian).
Tracy, I agree that I used terms vaguely, whereas they have a precise meaning in linguistics — just as splicing means one thing in the general domain, but something overlapping but different in molecular biology. Thinking over it yesterday, I came up with the same term as you did: vowel quality. This attribute overlaps with other “musical” qualities of speech.
Okay, folks, no need to start a flame war over linguistic terminology.
Especially since as far as I know, Juliette is the only actual trained linguist here.
As some of you noted, Athena’s point is well taken–there are often phonological differences that are clear to a native speaker but opaque to a non-native.
By the way, although English is not lexically tonal, it does use tonal to connote a lot of information, such as incredulity, certainty, and so on. It’s my understanding (also not being a trained linguist) that Chinese, which uses tonal lexically, actually has lexical particles that indicate some of the things that English does with tone.
I’m a trained linguist – I have an advanced degree and job experience, and I regularly do presentations at SF cons for non-linguists who are curious about language.
I am happy to take time to explain terminology from our field to people. I’m a lot less happy to be told this is flaming.
I’m with Tracy on this—I fail to see how pointing out that someone is demonstrably wrong, and suggesting that they take a look at some information to become better informed, counts as a “flame war.”
I’ll be finishing my PhD in linguistics next summer.
As to Chinese, I’m not sure if they have lexical markers for the things that we use intonation for in English, but this has nothing to do with Chinese being tonal; lots of languages have morphemes for this kind of thing.
Okay, than I remove my remark about not being a trained linguist. I run into too many autodidacts on “teh interweb,” who get fussy about things
.
However, the discussion was getting a bit testy. And a bit self-righteous all around. Just asking folks to take things a bit slow and calm… I would prefer the conversation keep to the friendly level. Doesn’t mean that correcting information can’t be passed on. But remember–English *does* use intonation to suggest a lot of things, which is completely lacking in web discussion, and that lack thereof can lead to misapprehensions. And to flame wars. Even if not here.
Jumping in here… There are a number of issues that have been brought up. While I’m not familiar with a vast number of languages, I do know that “tone” in the Chinese context (and I think probably beyond Chinese as well) is used to describe the way that Chinese uses specific intonational contours to differentiate between syllables with the same vowel. While this is an example of intonation, it’s a very specialized use of same. English word-level intonation (not sentence-level) is generally described in terms of “stress” while Japanese intonation is called “pitch accent.” Each of these works very differently even though they utilize some of the same basic vocal mechanisms.
The distinction between cup and cap is one of tongue position (vowel quality, as Chris and Tracy said), but the question of distinguishing the two can be complicated by variants in English dialects. Australian “u” in “cup” is quite similar to American (Californian!) “a” in “cap,” for example, though the two are more clearly distinguished in my own dialect of American English.
I think one of the points that I wanted to get across in my original discussion is that precise linguistic terminology should not become a barrier to people using linguistic principles in their fiction. On the other hand, a bit more research into the precise terminology is warranted if you have a character who is a linguist and needs to use such words as “intonation” or “tone” etc, so that the person you are placing in a position of expert can enact his or her role appropriately.
I appreciate everyone’s contributions, and how deeply they care about this topic.
This is just a clarification about tone, not a disagreement about the larger point.
Chinese does associate pitch contours with individual syllables. There are other ways tone can work – for example, some languages associate the tone “melody” with an entire word, meaning that when other rules shorten or lengthen the word, the melody may end up stretched to cover more syllables, or compressed to cover fewer.
Very informally, if I wanted to cover tone languages in general, I would say that “tone” means that a change in pitch can change the meaning of a word.
Some people have mentioned pitch in English. English has some rules that deal with pitch – for example, the difference between “You’re going to the store” and “You’re going to the store?” is conveyed by pitch – but the individual words don’t change meaning because their melody changes.
Tracy, thanks for your additional point. I think you’ve brought up something important, which is that the central issue is where (which linguistic area) tone is exercising its effect. Tone in Chinese is operating on the level of phonology, creating minimal pairs in which the meaning of a single-syllable word is entirely altered by a change of tone. The tone melody example is on the whole word level. Japanese pitch accent is assigned on the level of mora (segments of a syllable) where *ha shi de su (it’s chopsticks) is different from ha *shi de su (it’s a bridge) which is different from ha *shi *de su (it’s an edge). [please forgive the notation, which is not standard - I don't know how to do the actual notation in a blog comment!] A similar pitch change is operating at the sentence level in English to distinguish questions from statements. I’ll look forward to talking about these different levels of complexity as the series goes on.
[...] in character, a post that I found very enlightening. She also points the way to the “How Linguistics Can Help You” series at the Science in my Fiction blog. Thanks, [...]
The beauty of non-human language is that it needn’t be dependent on our physical limitations. I think about dolphins and elephants and cuttlefish in particular, although I’m sure I could start a fight by declaring their complex communications ‘language’.
But what’s important is that their communications are totally shaped by physical features humans lack. So when we create an alien language, we can apply the principles of human linguistics (there must be patterns, right?) but we don’t have to give them specifically human limitations.
Of course, if your aliens whistle, flicker, or chime, there will be trouble transcribing it. An opportunity for multimedia publishing? Or a way for weak writers to cop-out of doing solid worldbuilding?
One of the best stories on this topic is Ted Chiang’s The Story of Your Life. It touches on all the basic problems: transcription, the importance of a “Rosetta Stone”, the difficulty with relative values (or fine graining) attached to concepts — and the fundamental difference between spoken and written language in terms of both coding and decoding. The two are often conflated in casual conversation, but the distinction becomes glaringly obvious when dealing with spoken languages that adopted scripts poorly suited to their phonemes (hanji/kanji for Japanese; arabic script for Turkish).
Athena Andreadis revealed:
“I am not a professional linguist”
ÁLL I nèeded to knów.
(In Chinese, Chinese characters are hanzi, btw. Hanji is Korean paper.)
Juliette, these posts are fun. (Saw the link to the morphology one on Twitter this morning, dropped by to read both.) How many do you plan to do in the series? (One a month only?)
(from Sunshine Iwaki, Japan)
I didn’t make both substitutions from kanji to hanzi. An error — though the Latin choices are quasi-arbitrary, given that both renditions are approximations of the original phonemes.