A crash course in Cognitive Grammar David Tuggy SIL-Mexico Bio etc. A second-generation “missionary kid” from Venezuela. Grew up with two languages, linguistically-aware family. Have worked in Mexico with Nahuatl, 2 ling/trans programs. Fortunate to work on/with CG since its earliest days. www.sil.org/~tuggyd What to expect CG is a theory of linguistics, not of translation. It helps me understand what language is / languages are like. It seems difficult, if not in principle impossible, to see clearly what translation is before you have a good grasp of what languages are. What to expect CG has helped me synthesize Everything I know to be true about languages fits into it pretty well—seamlessly in most instances. My hope is that it will help you as well. Then you will understand *why* translation is so hard. (But don’t expect it to make translation easy. It may even make you more aware of how complex it is.) What to expect Some parts of this presentation are cobbled together from previous presentations. Some minor inconsistencies, a bit of incoherence, a fair amount of repetition (not necessarily a bad thing.) If it’s too elementary for some of you at times, too bad. I want to make sure you are with me. Hollenbach’s principle. What to expect Rules: Somebody please keep time on me, and interrupt me every 15 minutes so I can ask “Are you all on board with me?” Feel free to ask clarification questions. Will try to make time for more in-depth discussion too. But lots to present, so will often stick pretty closely to ppt presentation. What to expect I tend to talk too fast when I get going—slow me down. I’ve made no handouts. I’m happy to give you the power points. For those who read Spanish, a lot of this stuff is in a book (under development) which you can ask for electronic copies of. david_tuggy@sil.org Two ways with linguistic theories Look at them. (Legitimate.) Look through them. (Potentially more useful.) Cognitive grammar Is a theory developed largely by Ronald W. Langacker. It is part of a larger movement known as “Cognitive linguistics” One of its basic premises is that language is very closely linked to other cognition. I.e., how we talk is not terribly different in many respects from how we do lots of other things. What is a language? A language is a structured inventory of conventionalized linguistic units Write this definition down, and memorize it. All of the major words in it are technical terms. Starting with “unit”, then: A language is a structured inventory of conventionalized linguistic units A unit is a cognitive routine that you have mastered. You can activate it and run through it without special “constructive effort”. “Unit formation” or “entrenchment” is close to the ideas of “internalization” and “habit formation”. Our mental life largely consists in activation of one “unit” after, or concurrently with, another. A language is a structured inventory of conventionalized linguistic units Examples of units are walking, whistling “Dixie”, opening the refrigerator, stopping at a stop sign, reading the time from a clock, recognizing your cat, scratching your ear, figuring out what day of the week tomorrow is, pronouncing [ü], …(ad infinitum). Several characteristics of units should be obvious from the above examples. A language is a structured inventory of conventionalized linguistic units Some units are basically perceptual (e.g. recognizing your cat or a stop-sign) Others are basically motor-related (e.g. extending your leg or flexing your foot) Others are more “autonomously cognitive”; less directly connected to the perceptive/motor “periphery” (e.g. calculating how many days till next Tuesday.) A language is a structured inventory of conventionalized linguistic units Units are astonishingly flexible. You walk, without thinking about it, over all kinds of unpredictable terrain. You can open the fridge from all kinds of angles, with all kinds of different things in your hands, even a fridge you’ve never used before. You recognize your cat from virtually any angle, even when you only catch interrupted glimpses of it, etc. A language is a structured inventory of conventionalized linguistic units This flexibility may take the form of special versions of the routines in question. Climbing stairs, or running, may be seen as special kinds of walking Skipping may be seen as a special kind of running. But they can also be rightly seen as different from standard or prototypical walking or running. A particular action may exemplify more than one routine: it makes little sense to worry over whether a person running upstairs is running or climbing stairs. A language is a structured inventory of conventionalized linguistic units Units are often very complex. There are layers upon layers of units; units recruit other units as part of their activation. Moving your legs cyclically, keeping the right —constantly changing— amounts of muscular tension on your feet, ankles and thighs, belly and back, keeping your torso upright, interpreting cochlear and other signals regarding your balance and adjusting accordingly, interpreting visual signals as to your current position and where you expect to go, adjusting for obstacles, all are part of walking. Recognizing a stop sign, extending your leg to the brake pedal (without looking!), pressing it the right amount for the right length of time, checking for traffic, starting forward again, all are part of stopping at a stop sign. A language is a structured inventory of conventionalized linguistic units A single unit can be a part of many other units. Extending your leg is a part of walking and of stopping at a stop sign, of kicking a soccer ball and of stretching or …. Walking is usually part of opening the fridge door; it is also part of carrying a plate to the dinner table, or going to the front door or climbing the stairs or… Puckering your lips is part of (certain kinds of) whistling, and of kissing, and of pronouncing /w/, and of … A language is a structured inventory of conventionalized linguistic units A unit is typically more than the sum of its parts. Knowing how to walk and how to juggle is not the same as knowing how to juggle as you walk. Knowing how to hold your tongue to pronounce [i] and how to round your lips for [u] is by no means the same as being able to pronounce [ü]. Knowing how a pawn can move two squares on the first turn, and how it captures diagonally, is not the same as knowing how to capture en passant. A language is a structured inventory of conventionalized linguistic units Units’ effective simplicity (you can “wield” them as wholes) does not necessarily (or even usually) mean you forget their complexity. We don’t forget (though we may rarely pay attention to it) that walking involves putting one foot in front of the other. We know that serving a tennis ball involves throwing it up with one hand and hitting it with a racket in the other, and that the ball is supposed to go over the net and into the proper “box” on the court. A language is a structured inventory of conventionalized linguistic units Many units are culture-related (e.g. whistling “Dixie” as opposed the William Tell Overture, recognizing the shape of a stop-sign, assessing whether a complete stop is necessary or when it’s safe to go, serving a tennis ball, making tortillas the Mixtec way or starting a fire the Nambicuara way.) A language is a structured inventory of conventionalized linguistic units Units are established by usage. You learn how to walk by doing it (and to some extent by watching other people do it). It is not simple at first, but takes a lot of concentration (“constructive effort”). With long practice it becomes easy to the point of being almost automatic. Tying your shoes or reading your name the first time is a great accomplishment; with long practice you can do it unthinkingly. A language is a structured inventory of conventionalized linguistic units The number and frequency of usages is important for entrenchment. It typically takes many repetitions to thoroughly entrench a unit. High-frequency routines tend to be more central to language (and other parts of our lives). Salience (“emphasis”, cognitive energy) of a usage enhances the entrenchment process. You are likely to learn faster and remember longer how to recognize what you are afraid of, or to master something you find fun(ny) or beautiful. A language is a structured inventory of conventionalized linguistic units Units are routines: they are in some basic sense procedural. Yet in many, including many perception-related ones, the result of the procedure may be a particular (very complex) state of activation (salience) of various parts of the mind. When we refer to linguistic structures we are in fact reifying those procedures and/or the cognitive configurations they produce. A language is a structured inventory of conventionalized linguistic units Unit status is a matter of degree. Routines may be more or less thoroughly entrenched or mastered. The entrenchment process starts with the first usage, but it typically takes many more usages before the unit is firmly established. Some of the units that constitute a language are more thoroughly mastered than others. A corollary is that there is no sharp dividing line between what is “in” a language and what is not. Structures are constantly on their way in (being learned) or out (being forgotten). A language is a structured inventory of conventionalized linguistic units Language structures are a subset of the enormously complex set of units that people learn. They are different from other units in that they are linguistic. For Langacker, the prototypical core of language is the (essentially Saussurean) sign: A signifiant (prototypically a phonological structure) is strongly associated with (has a symbolic link to) a signifiée (semantic) structure. A language is a structured inventory of conventionalized linguistic units There are, then, three (and CG asserts, only three) basic kinds of units in a language: Semantic units Phonological (or other signifiant) units, and Symbolic units I.e. “bipolar” units consisting in the pairing of a semantic “pole” with a phonological “pole”. There are no separate syntactic or pragmatic units (etc.) Syntactic units are a particular kind of symbolic units. Pragmatic structures that achieve unit status are semantic. A language is a structured inventory of conventionalized linguistic units The linguistic units that make up a language are conventionalized: That is, they are conventional: shared and known to be shared by the relevant group of people, and Their conventionality has unit status, i.e. it has been established by communicative usage. A word or phrase that I make up and use privately is not part of a language until I and someone else use it and successfully communicate with it. A language is a structured inventory of conventionalized linguistic units Even then it isn’t part of the language we share with others until they learn and use it as well. This criterion, then, introduces a number of parameters along which gradations between what is centrally part of a language and what is clearly not may be observed How strongly conventionalized is the routine For what group(s) of people In whose understanding of it (speaker and hearer may differ) In what social contexts … A language is a structured inventory of conventionalized linguistic units A language is called (and thought of as) an inventory to contrast with a couple of other, flawed, notions of what a language is. A language is not a constructive device. The language does not generate sentences. People do. The language is their toolbox, the set of procedures they use to generate and to understand sentences and (more importantly) to talk to each other whether in sentences, isolated words, phrases, paragraphs or whatever. A language is a structured inventory of conventionalized linguistic units A language is not a sleek, streamlined piece of computer programming that automatically and elegantly produces myriad astounding results —all and only the correct ones, of course— with a few simple statements. Much of the serious description of any language looks more like a bunch of laundry lists. A language is a structured inventory of conventionalized linguistic units It is a huge set of huge lists. There are very many overlapping, redundant, quirky, idiosyncratic, iffy structures that show up as part of a language. I.e. they are (in some degree) units, conventionalized, and linguistic regardless of how nearly predictable they may be. Being almost completely predictable does not mean they are excluded. Being idiosyncratic does not mean they should be ignored. A language is a structured inventory of conventionalized linguistic units This inventory is not haphazard, however. It is densely structured. Much of the structuring is through the inclusion relationships we described for units generally: many linguistic structures recruit others as parts of them. So component-composite relationships are an important structuring mechanism. Associations of one component with another (from participation in a composite) are also very important. Symbolic linkages are such a kind of associations. A language is a structured inventory of conventionalized linguistic units The inventory is also saliently structured because some units categorize others. We will be talking a lot about this later: these are schematic and partially-schematic relationships What are traditionally called “rules” are understood in CG to be schematic structures which categorize more specific structures. Ultimately all these relationships, composition, association and categorization, are intimately related, not totally different in kind. In the next powerpoint file, we will start again from the beginning and cover some of the same ground again, presenting it from a different (though perfectly compatible) perspective.