It's all relative (part 2)

In my last post, I gave a brief potted history and explanation of the concept of linguistic relativity. Now I want to touch on some of the reaction to this new wave of research on linguistic relativity and talk a bit about my own take on it.

Through most of the 20th century there really wasn’t much solid evidence supporting the notion of linguistic relativity, and as I’ve said linguists were mainly focussed on looking for universals in language. So it is perhaps not surprising that in spite of the growing body of experimental evidence to support the idea, there is still considerable resistance. But in the recently burgeoning field of cognitive linguistics, linguistic relativity is starting to gain real traction again.

As Daniel Casasanto points out, most people who argue against the idea of linguistic relativity are objecting to the wrong thing.1 Most proponents of linguistic relativity don’t hold that language is the same thing as thought, just that it can influence thought. Casasanto proposes the simple (and quite elegant) mechanism of learning. Through repetition, language trains us into thinking in certain ways. Learn a new language and you are gradually trained into a new way of thinking. This is a convincing explanation that seems to fit much of the the data well. There is also evidence to suggest that we sometimes draw on language as a tool to aid thought. In some experiments, when your ability to use language is negated (by tying up your language faculties with other linguistic tasks like reciting a series of numbers), the linguistic relativistic effects lessen or disappear. Thus there may be more than one mechanism at work here. Clearly there is still much work that remains to be done to figure out the exact scope and parameters of this Whorfian effect.

It's interesting to note that many of the people who argue against linguistic relativity tend to focus only on strong linguistic determinism or other notions, such as the idea that language is thought, that few if any still argue for, and largely ignore the more recent work in this area. It's a bit of a straw man argument really. Or they'll focus solely on the false claims made in popular media on the topic. See for instance this very revealing debate on linguistic relativity from a few years ago. There certainly are a great deal of misleading and downright incorrect claims popularly made about linguistic relativity which is quite rightly deserving of criticism, but the popular media has a long history of misrepresenting the ideas of scientists and other academics, and none of this really reflective of the merits of the real argument itself. I do think that the weight of opinion will gradually shift as more research is done. I'm certainly convinced that at least some of what is claimed about linguistic relativity is true and valid, and well supported by the evidence that I’ve seen, but much of this research is still on the cutting edge of the fields of linguistics and cognitive science, so time will tell.

For my own part, it seems to me that we frequently ask the wrong questions when debating the validity of linguistic relativity. Does language influence thought, or does thought influence language? Well it’s likely both. And clearly there are external factors on both, what we might refer to as culture.2 It is perhaps best to think of a complex three-way relationship between language, thought, and culture, in which each influences the other two, which may lead to feedback loops with language, thought, and culture reinforcing each other back and forth. Furthermore, it is not at all surprising that often real-world evidence of Whorfian effects seem hazy, as one would expect multiple, often conflicting, linguistic influences on thought as well as multiple cultural influence. Rarely if ever would it be the case that an element of language is the only thing that influences some aspect of the way one thinks or behaves. But just because the Whorfian effect is frequently cancelled out or overruled doesn't mean it isn't operating beneath the surface. In highly controlled experimental situations, such as those Boroditsky creates, it is often possible to isolate a Whorfian effect, but it is such a complex system that practically speaking these effects may be hard to see clearly in the real world, though sometimes it seems to come out in startling ways. Again, even if the effects aren’t clearly visible, it doesn’t mean that the influences are not there. Human behaviour is complex and not easily reducible or predictable. See also my previous comments on interconnectivity -- it's the complex nature of the connections that's really interesting here.

I'm not myself concerned with isolating these influences or devising some kind of predictive theory about behaviour, though those lines of research are certainly important themselves. Since I come more from the humanities side of things rather than the sciences or social sciences, I'm more concerned with interpreting rather than predicting, so my main interest is seeing what linguistic relativity can tell us about the complex interrelationship between language, thought, and culture over time through the course of history. And as I've said, I see all these elements influencing each other. The (admittedly imperfect analogy) I like to think of is that even though the gravitational effect of a pebble on the earth is tiny compared to the effect of the earth on the pebble,that doesn't mean that it isn't there, and if enough pebbles start pulling in the same direction, metaphorically speaking, the effects can add up. Of course the tendencies of individuals may in many cases be far more variable than any Whorfian effect. The Whorfian effects might be the tipping point in an otherwise evenly balanced situation, or might only be noticeable in isolation or in the aggregate; but if they line up with other cultural influences the outcome might be profound, with the kind of feedback loop effect I alluded to earlier.

In any case, I've waffled on about this topic no doubt to the point of boring any casual reader into giving up on this post several paragraghs ago, so I'll stop now, though it's likely that the topic will come up again in future posts.

1 See Casasanto's excellent and wonderfully titled article "Who’s Afraid of the Big Bad Whorf? Crosslinguistic Differences in Temporal Language and Thought". [back]

2 See Dan Everett’s fascinating work on the cultural constraints on language. Based on his research on the fascinatingly different language Piraha, Everett rejects the notion of universal grammar and concludes that language is constrained by culture and is in fact a cultural tool, an invention by human beings, not an innate instinct. See also Everett's more popularly targeted book Language: The Cultural Tool. I'll be writing more about these topics in the future, particularly as they relate to time in language and thought. [back]

It's all relative (part 1)

This is the last of the background posts I had planned (for the time being anyway). In the others, I have written about cognitive philology, interdisciplinarity, and interconnectivity. Here I wish to briefly state my position on linguistic relativity. I’ve touched on the topic before in a number of other posts, and it was the focus of a conference paper I mentioned, but I thought it was worth briefly summarising in one post what exactly linguistic relativity means (to me in any case), and what position I take on this often controversial topic. I’ll break it into two posts to keep it from being too long, so expect the second part to follow shortly. I’ve briefly explained linguistic relativity before, but here in more detail is the summary I originally wrote (in slightly modified form) for a conference paper, with special focus on some recent research on the topic by Lera Boroditsky.

The idea of a relationship between language and thought was touched on to some extent by such early 20th century anthropologists as Franz Boas and Edward Sapir, but is was really due to the systematic treatment by Benjamin Lee Whorf, a student of Sapir, that it became a focus of attention. Though Sapir himself never directly argued for a strong causal link between language and thought, Whorf’s bold claims about such a connection have led to the idea being often referred to as the Sapir-Whorf hypothesis. Simply put, the Sapir-Whorf hypothesis proposes that differences in linguistic features (and that could be anything from words to the kinds of grammatical information expressed in a language, like verb tense, to syntactical structures used) from one language to the next determine or at least influence the ways in which speakers of those languages think about the world. For instance, Whorf made the observation (largely incorrectly as it turns out) that “the Hopi language is seen to contain no words, grammatical forms, constructions or expressions that refer directly to what we call ‘time,’ or to past, present, or future” (57) and thus he concluded that it was “gratuitous to assume that Hopi thinking contains any such notion as the supposed intuitively felt flowing of ‘time’” (58).1 There have come to be two versions of this hypothesis, the strong and weak versions, sometimes distinguished as linguistic determinism and linguistic relativity. The strong determinism version maintains that the particular linguistic categories available to a speaker determine and constrain what the speaker is able to think, whereas the weak relativity version states only that language differences can lead to differences in thought. While few scholars still hold to the strong formulation of the Sapir-Whorf hypothesis, the weak version has remained an open question.

For much of the latter half of the 20th century, this notion fell out of favour and has been strongly rejected, especially after Noam Chomsky first proposed the notion of a universal grammar, which has since become the mainstream position. Chomsky specifically rejected Whorf’s idea of linguistic determinism, and instead held that language differences were surface-level only, and that all humans shared a common grammar driven by a common faculty of language acquisition. More recently, Steven Pinker states unequivocally that the theory is “wrong, all wrong”, and instead argues for an underlying language common to all people, which he calls ‘mentalese’, and further proposes that humans possess an innate instinct for language which has evolved through natural selection.2 Indeed the mainstream view for the past half century has been to search for universals in human language, and to assume that human cognition too is universal regardless of linguistic differences. As Whorf’s linguistic observations had fallen through, and in the absence of much real evidence in support of the Sapir-Whorf hypothesis, the notion of linguistic relativity remained either a fringe view or one of mistaken popular conception only.

However, in more recent years, perhaps only the past decade or a little more, many of these mainstream notions in linguistics have been challenged. For one thing, linguists are more and more starting to challenge the idea of a universal grammar,3 especially in the light of cognitive linguistics which seeks to explain linguistic phenomena on the basis of general cognitive principles rather than specialised language mechanisms. Furthermore, new experimental evidence in support of linguistic relativity is being generated by a number of cognitive scientists, most notably Lera Boroditsky, along with her various collaborators and other researchers working along similar lines.

Boroditsky’s research is based mainly on experimental procedures, often involving cross-linguistic evidence, generated by conducting controlled lab experiments on native speakers and measured by very quantitative results such as response times. She has conducted research on such phenomena as colour perception, demonstrating that possessing particular colour categories in one’s language improves one’s performance in non-linguistic colour discrimination tasks. Thus Russian speakers can more quickly distinguish between two distinct and mutually exclusive shades of blue, goluboy ‘light blue’ and siniy ‘dark blue’, than English speakers, who collapse all such shades under the one umbrella term blue. Boroditsky has also shown the effect of agentive language on eye-witness memory, with speakers of languages that grammatically require an agent even for accidental events having better memory of the agent involved in an action which they have witnessed; this is the case with English, which uses more agentive language, as opposed to Japanese or Spanish, which use less agentive language, for instance. The quantity of agentive language also correlates with the assignment of blame and even harshness of penalties, both in laboratory experimentation and in analysis of court transcripts. Boroditsky has also demonstrated that grammatical gender has an impact on how we think about objects, assigning gendered qualities to them and influencing how abstract concepts are allegorised in art, a result that will be of little surprise to most art historians.

Boroditsky has also done a considerable amount of work on time cognition, particularly the use of spatial metaphor for describing and reasoning about time. She has demonstrated, for instance, that different languages use different spatial metaphors, front/back in English, and up/down in Mandarin. It turns out that people use these varying spatial-temporal metaphors even in non-linguistic cognitive tasks, and thus think about time in different ways. Also in terms of spatiotemporal metaphor, one can think of either time moving, as if you are watching a river flow towards you, for instance, as in “the holidays are approaching”, or ego-moving, as if you yourself are moving along a path “we’re rapidly coming to the end of term”. And it turns out you draw on spatial reasoning actively, and being primed with one spatial arrangement or another can affect your performance of temporal reasoning. Furthermore, whether you conceive of time as moving left to right, as English speakers do, or right to left, as Hebrew speakers do, seems to be dependent on the writing direction in your language; as well, different languages conceive of the future either in the front, as in English, or behind, as in Aymara. And in reversing the perspective, ego-moving or time moving, we also reverse the directionality, such as front/back. Thus, in ambiguous statements such as “We moved the meeting forward a day”, forward can be taken to mean more in the future or less in the future. Perhaps most striking are the differences between languages which use body-relative indications of space, as in right and left in English, and languages which use only absolute directions such as cardinal directions (north, south, east, west). The Pormpuraaw languages of Australia, such as Kuuk Thaayorre, are such languages. Speakers of these languages are always aware of their absolute spatial orientation, and furthermore draw on this spatial reasoning for reasoning about time, always arranging pictures in temporal order east to west regardless of the orientation of their own body. Differences in metaphors used for duration of time are also cognitively significant, such as with linear distance in English as in “a long time” and amount in Greek “poli ora” or “much time”.

With many of these effects, Boroditsky has demonstrated that not only does the brain seem to use language online for non-linguistic tasks, which can be demonstrated by running language interference tasks which disable one’s ability to use language in the moment, but also that the effects of language differences are also long-term. Thus bilingual speakers perform differently depending on the linguistic context – they essentially think differently depending on which language they are using. But also, learning a new language can change your cognitive performance even when you are currently using your old language, and can even change how you speak your own language, and thus speakers of languages in which tense-marking is optional have been shown to mark for tense more often if they speak another language which requires tense marking. Of course Lera Boroditsky is not the only scholar working on the question of linguistic relativity, but I mention her work in detail since much of it focusses on language and temporal reasoning, which is of particular interest to me.4

In the next post I’ll touch on some of the more recent reactions to this research and talk a bit about my own take on it.

1 Benjamin Lee Whorf, Language, Thought & Reality (MIT Press, 1956). [back]

2 Pinker's book The Language Instinct was so persuasive to me, in spite of the fact that I now think he is in many ways fundamentally wrong on this, that I think I was rather too timid in the position I took in my doctoral dissertation. [back]

3 A very broad debate that I can’t go into here but see this article for a thorough treatment of the topic. [back]

4 There is of course much I’ve left out here, such as the “thinking for speaking” model of Dan Slobin, but I wanted to focus mainly on the aspects that are most relevant to my own interests. If you want more information on this topic, you would also be well advised to read Boroditsky’s entry on linguistic relativity in the Encyclopedia of Cognitive Science, as well as this more detailed and somewhat more critical account from the Stanford Encyclopedia of Philosophy. In future posts I’ll also go into a bit more detail about Boroditsky’s work on time which I’ve summarised above. [back]

Time off and time on

As is no doubt obvious, this blog has gone dormant for a time due to the busy-ness of life, what with kids home for the summer and then a busy academic year. But after this unfortunately extended time off from this blog, I want to pick up where I left off.

As promised ages ago, I first want to write (or rather finish writing) a post that both explains and lays out my position on the the concept of linguistic relativity, and how I think language, thought, and culture are interrelated. I also have few one-off ideas on topics such as semantics, memetics, and the formulaic nature of language, as well as some hopefully fun and light-hearted posts (which I'll keep under wraps for the moment).

But my big plan is to go back to the topic which I've been working on, on and off, for over 15 years now, time in language and thought. My doctoral dissertation was on the coneptualisation of futurity in Old English, and came out of an interest in how English developed from a language that didn't originally have a systematic way of expressing the future, and how this might relate to the Latin language which does have grammatical future tense, the adoption of a new religion, Christianity, which has important conceptual foundations involving a notion of future time, new philosophical concepts such as the free will/predestination debate, etc. So first of all, I'll review some of that research (and how my thinking may have changed in some ways since then), but also much of the attendant background to this topic, about time in general in thought and language, which will give me a chance to write about some of the stuff that was too tangential or just didn't fit with the dissertation.

There has also been an explosion of work over the past decade or so, and particularly in the last couple of years, on topics that relate to this, such as the work by Lera Boroditsky and a number of other scholars on how we use space to think about and talk about time (and also in particular the research by Daniel Casasanto and Vyvyan Evans), or by David Eagleman on time perception, and even popular nonfiction books on time, such as Claudia Hammond's Time Warped and physicist Adam Frank's About Time. My own most recent conference paper dealt in large part with spatio-temporal metaphor in Old English. But in an attempt to keep these posts fairly informal, easily approachable, and not overly wrought, as was my original intention when I started this blog, I'm going to try to write about these topics in an off-the-cuff manner without putting a lot of advanced planning into organization. So they'll be a little haphazard as I jump from one topic to another as I think of them in an attempt to get the ideas down quickly.1

It's often pointed out that time is the most common noun in the English language (and is similarly high ranking in other languages too), and that three of the top five nouns are time-related words (the other two being year and day). Personally, I've always been interested in and maybe even obsessed with the notion of time, how we think about time, how we talk about time. I've always been fascinated by clocks and watches,2 and I've always been drawn to stories about time and especially time travel, such as the tv show Doctor Who and my favorite children's book The Story of the Amulet by E. Nesbit. Even before I started in on my graduate school work I became fascinated in the free will / predestination debate and as an undergraduate read Boethius' Consolation of Philosophy and the debate on free will between Desiderius Erasmus and Martin Luther. Indeed even now I often include a number of texts in my regular teaching rotation that pick up on themes and motifs of time, such as Alice Munro's "Walker Brothers Cowboy", Stephen Leacock's "The Retroactive Existence of Mr Juggins", Robert Heinlein's "By His Bootstraps", and the Christopher Nolan film Memento, so I also want to post about literary representations of time, including more popular forms of literature and entertainment.

So if you're interested in a somewhat chaotic series of reflections on time in language, thought, and culture, then watch this space...3

1 Organizing these ideas into something more cohesive may be a project for a later time. [back]

2 An antique Waltham pocket watch given to me by my mother is one of my prized possessions. [back]

3 I’m hoping that having posted this, I will now be more motivated to follow through with it and do it, but any words of encouragement will certainly help give me a kick in the complacency. [back]