Knitting and linguistics

So.  I think I said in my last post that, in our conversation during her class, Cat described to me the way she's been thinking about knitting in terms of linguistics.  Needless to say, I loved it.  So (because I like to share the things I love), I thought I'd try to explain the whole thing here, and see what you all think.  This means that there is going to be a linguistics lesson.  This is, in fact, a very fibery sort of post, in the sense that (if it all goes well), we'll be talking about knitting by the end, and talking about it analytically, but if linguistics isn't your thing, I completely understand.  Next time, there will be pictures of knitted objects and spun yarn (yes, I have been spinning!).  Note, you will need to say things out loud during the reading of this blog post - shut your office door, if necessary.

I'm going to lay out what we said in class in brief, just to kick things off, and then back up and define the main linguistic terms.  Then I'll go at the linguistics/knitting thing again, except it will be clearer, and you can (and should!) join in the fun. (I should say that there is just no way to do this justice in one blog post - this is just sort of a kick-off.)  In essence, Cat said that she was intrigued by the linguistic unit of the "phoneme", and thought that it could apply to knitting.  I even wrote to Cat to make sure that I remembered what she said about how that would work (by the way, can I just tell you how much fun it is to meet someone who's as enthusiastic about thinking this way as I could ever hope to be, and who is totally willing to talk to me about it, even though I'm sure she's crazy busy?): "Any unit of stitch manipulation that could be placed into or removed from a piece of knitting to alter its appearance/ texture. This would include single-stitch units such as knit, purl, yo, etc., as well as larger horizontal units like cables and larger vertical units like slipped stitches, etc.  What seems essential is that each unit, whether simple or complex, carry its own distinct identity/meaning (thus morpheme is better than phoneme)."  That last bit is because when she expressed this idea to another linguist in another class, that person suggested that the idea of "morpheme" might better capture what she was thinking.

And so beginneth the lesson.  First, the "phoneme".  Phonemes are defined as units of sound in a particular language.   The main idea to get here is that speakers of particular languages treat sounds which are, acoustically, different from one another as if they were the same.  In doing so, they are acting out an unconscious agreement among the speakers of that language that those variants will not change the meanings of words if they are changed.  So, for example, in English, we use a sound which I will represent with the symbol [t] (a linguist would say that this symbol stands for a sound which is described as a "voiceless alveolar stop").  You can hear that sound in a word like "stop"; it is the second sound in that word.  (I won't get into it in too much detail here, but it is absolutely critical to remember that spelling is NOT language, and that we are talking here about SOUNDS, not about the way those sounds are represented alphabetically.)  There is another sound in English that is very similar - it is represented by linguists as a [t] with a little superscript "h", which I can't make Blogger do, so we'll call it "aspirated t", which is what it is.  It is very much like the first [t], except that it has a little puff of air associated with it - you can feel that if you say the word "top".  Feel how there's a much bigger puff of air when you say the [t] in "top" than when you say it in "stop"?

Now.  You tell me.  Would you consider those two sounds (and believe me, acoustically, they are definitely two sounds) to be two different sounds, or simply variants of the same sound?  In other words, are they as different as [d] and [t], or not?  Note that [d] and [t] are "different" enough to speakers of English that they will distinguish between words - for example "dill" and "till".  But would [t] and aspirated [t] do that?  

Most native English speakers will, at this point, either say "no", or they'll look at me like I'm nuts for even asking ("duh" being a stronger version of "no").  For English speakers [t] and aspirated [t] are simply variants of the "same" sound, while [t] and [d] are "different" sounds.  (But make the sounds [t] and [d] for a second, paying very close attention, and notice how similar they are.  Acoustically and articulatorily - the way you move your mouth - they are as similar as [t] and aspirated [t] are; [t] is a voiceless alveolar stop, and [d] is a voiced alveolar stop - the only difference between them is whether your vocal cords are vibrating or not.  Cool, huh?)  This is specific to English - some languages view [t] and aspirated [t] as being utterly distinct sounds, as different as [t] and [d] and just as able to distinguish between words (Bengali is an example of one such language).  But for English, the sounds [t] and aspirated [t] are variants of what we consider to be "one sound" - linguists call that "one sound" a phoneme, and we call its variants "allophones".

One way to think about it is this.  Clark Kent and Superman are really the same guy, right?  I mean, they look really different from one another, and they always show up in entirely different environments, but they're variants of one person - [t] and aspirated [t] are the same way.  When he's at home, alone, with no-one to see him, who is this Clark Kent/Superman guy, anyway?  He's the phoneme.  And Clark Kent and Superman are his allophones.  They are variants whose appearance is conditioned by environments - Clark Kent writes newspaper articles and pines after Lois Lane, and Superman gets to save the world.  In the same way, aspirated [t] appears by itself at the beginning of words, and [t] appears in other places in words.  (It's more complicated than that, but I'm simplifying.)

(For another fun exercise, say these pairs of words: beet/bead; etch/edge; bus/buzz; block/blog.  Do you notice that in each pair, there is one word where the vowel in the middle is just a little longer than in the other word in the pair?  Do you notice it's always the second word that has the longer vowel?  Those vowels - short and long - don't distinguish between words in English (if you say "beeeet", you'll just sound like you're emphasizing the fact that you're talking about beets rather than turnips), but they are distinct from one another.  Long and short vowels are allophones in English.  There are other languages - like Kawaiisu - where short and long vowels are distinctive phonemes in the language - they will distinguish between words.)

Phonemes have no meaning by themselves.  But by contrasting with one another, they can create meaning within words - no English speaker will confuse words like "till" and "dill".  So, phonemes are minimal sound units within a particular language.  They are building blocks, but by themselves they mean nothing.  And phonemes have variant forms which cannot create meaning distinctions.

Now we move to morphemes.  Morphemes are defined in linguistics as "minimal meaning units" - the smallest bits of language that have meaning.  In a word like "cats", there are two morphemes - the morpheme "cat", which means furry thing with whiskers that purrs, and the morpheme "-s", which means more than one of whatever it's attached to.  Morphemes, too, have variant forms (we call them "allomorphs").  In the case of the plural morpheme in English, there are three variant forms - pay very close attention to the way you pronounce the plural morpheme at the end of the words "cats",  "dogs", and "churches".  Really, really close attention.  If you do (and it can help if you put them in front of another word, like "are", in a sentence), you'll notice that in the first one, we pronounce the morpheme [s], in the second [z], and in the third, something like [ez] (for any linguists out there, I'm sorry - I can't make Blogger speak IPA).  Cool, huh?  But we wouldn't want to say that there are three different bits that mean "plural" - we'd want to say that these are just variants of the "same" plural morpheme - they are allomorphs of the plural morpheme.  (There are lots of other examples, but just by way of one more, think about words like "impossible", and "intangible" and "inconceivable" - if you pay close attention, you'll hear that you say that last one, in casual speech, more like "ingconceivable".  That prefix im/in/ing-, meaning "not so much", has three allomorphs.)

OK.  Back to knitting.  The question is, are there units of knitting that are comparable to phonemes and morphemes?  Are there units which are basic building blocks of the language of knitting, with variant forms, which don't have meaning in themselves?  And are there other units (again, with variant forms) which do have some kind of "meaning"?  Looking even more broadly, is there a knitting syntax?  A way of combining larger and larger units of knitting to make "sentences" and even "narratives"?

After much contemplation (probably more than I should admit to, given that it's not even been two weeks since Cat dropped this lovely little cognitive plaything on me), I think that the knit stitch is a phoneme.  I think it has two allophones: the regular knit stitch, and the twisted knit stitch.  I think that the knit stitch contrasts with another phoneme, the purl stitch.  I don't think that either a single knit stitch nor a single purl stitch has meaning in itself, but I do think that they contrast with one another in larger knitted structures.  I also tend to think that a yo is a phoneme.  And maybe decreases and increases are phonemes, and the different kinds of, say, decreases are allophones of one phoneme.  In other words, a k2tog and an ssk will not distinguish different meanings within a knitted structure, but either one will contrast with a kfb or a M1.

I think that things like k2,p2 would be a morpheme - the "ribbing" morpheme.  I think that it's a morpheme because when stitches combine like that, they have a meaning within a knitted structure.  I also think that that morpheme has allomorphs, like the k1,p1 allomorph, or the k2tbl, p2 allomorph.  I'm wondering whether cables are morphemes, and at what point is one cable different enough from another that they're two different meaning units?  And, are small cables surrounded by purl stitches simply allomorphs of the ribbing morpheme, or something else?

And should we be calling these something else?  Alloknits and allomotifs?  Or...?

I love this analogy. It's making me think about what kinds of "meanings" are recognizable in our knitting, and I do think that there is such a thing.  When we see someone else's knitted garments, we know what they're trying to do.  We may not always find whatever it is appealing, nor think that it's the best way to convey a particular "meaning", but it's recognizable.  This is very much the same way that a syntactically appropriate sentence of English may be appealing to some speakers, while others think, "I wouldn't say it that way."  But it is recognizable as English, and it is recognizable because it follows the rules of the language which are acquired, unconsciously, by all native English speakers (I do not speak here of eighth-grade grammar class rules, which aren't the rules that linguists are interested in).  I do think, though, that there comes a point where this analogy may break down.  The combination of morphemes into words, and words into sentences, is subject to a pretty stringent and inflexible (all things considered) set of rules.  But that doesn't seem to be the case in knitting, at least not when it comes to knitted motifs (although probably when it comes to bits of garments - for example, you wouldn't want to put a sock cuff on the toe).  

And maybe that's the best bit about knitting?  It's rule-governed, but not rule-bound.  I think I like that.