3 burning issues in the study of metaphor

I’m not sure how ‘burning’ these issues are as such but if they’re not, I’d propose that they deserve to have some kindling or other accelerant thrown on them.

1. What is the interaction between automatic metaphor processing and deliberate metaphor application?

Metaphors have always been an attractive subjects of study. But they have seen an explosionof interest since ‘Metaphors we live by’ by Lakoff and Johnson. In an almost Freudian turn, these previously seemingly superfluous baubles of language and mind, became central to how we think and speak. All of a sudden, it appeared that metaphors reveal something deeper about our mind that would otherwise remain hidden from view.

But our ability to construct and deconstruct metaphors was mostly left unexamined. But this happens ‘literally’ all the time. People test the limits of ‘metaphor’ through all kinds of dicoursive patterns. From, saying things like ‘X is more like Y’ to ‘X is actually Y’ or even ‘X is like Y because’.

How does this interact with the automatic, instantaneous and unconscious processing of language. (Let’s not forget that this is more common)

2. What is the relationship between the cognitive (conceptual) and textual metaphor?

Another way to pose this question is: What happens in text and cognition in between all the metaphors? Many approaches to the study of metaphor only focus on the metaphors they see. They seem to ignore all the text and thought in between the metaphorical. But, often, that is most of goes on.

With a bit of effort, metaphors can be seen everywhere but they are not the same kind of thing. ‘Time is money’, ‘stop wasting my time’, and ‘we spent some time together’ are all metaphorical and relying on the same conceptual metaphor of TIME IS A SOMETHING THAT CAN BE EXCHANGED. But they are clearly not doing the same job of work for the speaker and will be interpreted very differently by the listener.

But there’s even more at stake. Imagine a sentence like ‘Stop wasting my time. I could have been weeding my garden spending time with my children instead of listening to you.’ Obviously, the ‘wasting time’ plays a different role than in a sentence ‘Stop wasting my time. My time is money and when you waste my time, you waste my money.’ The coceptual underpinnings are the same, but way they can be marshalled into meaning is different.

Metaphor analysts are only too happy to ignore the context – which could often be most of the text. I propose that we need a better model for accounting for metaphor in use.

3. What are the different processes used to process figurative language

There are 2 broad schools of the psychology of metaphor. They are represented by the work of Sam Glucksberg and Raymond Gibbs. The difference between them can be summarised as ‘metaphor as polysemy’ vs ‘metaphor as cognition’. Metaphor, according to the first, is only a kind of additional meaning, words or phrases have. While the second approach sees it as a deep interconnected underpinning of our language and thought.

Personally, I’m much closer to the cognitive approach but it’s hard to deny that the experimental evidence is all over the place. The more I study metaphor, the more I’m convinced that we need a unified theory of metaphor processing that takes both approaches into account. But I don’t pretend I have a very clear idea of where to even start.

I think such a theory would also have to account for differences in how inviduals process metaphors. There are figurative language pathologies (e.g. gaps in ability to process metaphor is associated with autism). But clearly, there are also gradations in how well individuals can process metaphor.

Any one individual is also going to vary over time and specific instances in how much they are able  and/or willing to consider something to be metaphorical. Let’s take the example of ‘education is business’. Some people may not consider this to be a metaphor and will consider it a straightforward descriptive statement along the lines of ‘dolphins are mammals’. Others will treat it more or less propositionally but will dispute it on the grounds that ‘education is education’, and therefore clearly not business. But those same people may pursue some of the metaphorical mappings to bolster their arguments. E.g. ‘Education is business and therefore, teachers need to be more productive.’ or ‘Education is not business because schools cannot go bankcrupt’.

Bonus issue: What are the cognitive foundations shared by metaphor with the rest of language?

This is not really a burning issue for metaphor studies so much as it is one for linguistics. Specifically semantics and pragmatics but also syntax and lexicography.

If we think of metaphor as conceptual domain (frame) mapping, we find that this is fundamental to all of language. Our understanding of attributes and predicates relies on the same ability to project between 2 domains as does understanding metaphor. (Although, there seems to be some additional processing burden on novel metaphors).

Even seemingly simple predicates such as ‘is white’ or ‘is food’ require a projection between domains.

Compare:

  1. Our car is white.
  2. Milk chocolate is white.
  3. His hair is white.

Our ability to understand 1 – 3 requires that we map the domain of the ‘subject’ on to the domain of the ‘is white’ predicate. Chocolate is white through and through whereas cars are only white in certain parts (usually not tires). Hair, on the other hand, is white in different ways. And in fact, ‘is white’ can never be fully informative when it comes to hair because there are too many models. In fact, it is even possible to have opposite attributes mean the same thing. ‘Nazi holocaust’ and ‘Jewish holocaust’ are both use to label the same event (with similar frequency) and yet it is clear that they refer to one event. But this ‘clarity of meaning’ depends on projections between various domains. Some of these include ‘encyclopedic knowledge’. For instance, ‘Hungarian holocaust’ does not possess such clarity outside of specialist circles.

It appears that understanding simple predicates relies on the same processes as understanding metaphor does. What makes metaphor special then? Do we perhaps need to return to a more traditional view of metaphor as a rhetorical device but change the way we think about language?

That is what I’ve been doing in my thinking about language and metaphor but most linguistic theories treat these as unremarkable phenomena. This leads them to ignore some pretty fundamental things about language.

How to read ‘Women, Fire and Dangerous Things’: Guide to essential reading on human cognition

Note:

These are rough notes for a metaphor reading group, not a continuous narrative. Any comments, corrections or elaborations are welcome.

Why should you read WFDT?

Women, Fire, and Dangerous Things: What Categories Reveal About the Mind is still a significantly underappreciated and (despite its high citation count) not-enough-read book that has a lot to contribute to thinking about how the mind works.

I think it provides one of the most concise and explicit models for how to think about the mind and language from a cognitive perspective. I also find its argument against the still prevalent approach to language and the mind as essentially fixed objects very compelling.

The thing that has been particularly underused in subsequent scholarship is the concept of ‘ICMs’ or ‘Idealised Cognitive Models’ which both puts metaphor (for work on which Lakoff is most well known) in its rightful context but also outlines what we should look for when we think about things like frames, models, scripts, scenarios, etc. Using this concept would have avoided many undue simplifications in work in the social sciences and humanities.

Why this guide

Unfortunately, the concision and explicitness I extolled above is surrounded by hundreds of pages of arguments and elaborations that are often less well-thought out than the central thesis and have been a vector for criticism (I’ve responded to some of these in my review of Verena Haser’s book).

As somebody who translated the whole book into Czech and penned extensive commentary on its relevance to the structuralist linguistic tradition, I have perhaps spent more time with it than most people other than the author and his editors.

Which is why when people ask me whether to read it, I usually recommend an abreviated tour of the core argument with some selections depending on the individual’s interest.

Here are some of my suggestions.

Chapters everyone should read

Chapters 3, 4, 5, 6 – Core contribution of the book – Fundamental structuring principles of human cognition

These four chapters summarize what I think everybody who thinks about language, mind and society should know about how categories work. Even if it is not necessarily the last word on every (or any) aspect, it should be the starting point for inquiry.

All the key concepts (see below) are outlined here.

Preface and Chapter 1 – Outline of the whole argument and its implications

These brief chapters lay out succinctly and, I think very clearly, the overall argument of the book and its implications. This is where he outlines the core of the critique of objectivism which I think is very important (if itself open to criticism).

Chapter 2: Precursors

This is where he outlines the broader panoply of thinkers and research outcomes in recent intellectual history whose insights this books tries to systematise and take further.

The chapter takes up some of the key thinkers who have been critical of the established paradigm. Read it not necessarily for understanding them but for a way of thinking about their work in the context of this book.

Case studies

The case studies represent a large chunk of the book and few people will read all 3. But I think at least one of them should be part of any reading of the book. Most people will be drawn to number 1 on metaphor but I find that number 2 shows off the key concepts in most depth. It will require some focus and patience from non-linguists but I think is worth the effort.

Case study 3 is perhaps too linguistic (even though it introduces the important concept of constructions) for most non-linguist.

Key concepts

No matter how the book is read, these are the key concepts I think people should walk away with understanding.

Idealized Cognitive Models (also called Frames in Lakoff’s later work)

I don’t know of any more systematic treatment of how our conceptual system is structured than this. It is not necessarily the last word but should not be overlooked.

Radial Categories

When people talk about family resemblances they ignore the complexity of the conceptual work that goes into them. Radial categories give a good sense of that depth.

Schemas and rich images

While image schemas are still a bit controversial as actual cognitive constructs, Lakoff’s treatment of them alongside rich images shows the importance of both as heuristics to interpreting cognitive phenomena.

Objectivism vs Basic Realism

Although objectivism (nothing to do with Ayn Rand) is not a position taken by any practicing philosophers and feels a bit straw-manny, I find Lakoff’s outline of it eerily familiar as I read works across the humanities and social sciences, let alone philosophy. When people read the description, they should avoid dismissing it with ‘of course nobody thinks that’ and reflect on how many people approach problems of mind and language as if they did think that.

Prototype effects and basic-level categories

These concepts are not original to Lakoff but are essential to understanding the others.

Role of metaphor and metonymy

Lakoff is best known for his earlier work on metaphor (which is why figurative language is not a key concept in itself) but this book puts metaphor and metonymy in perspective of the broader cognition.

Embodiment and motivation

Embodiment is an idea thrown around a lot these days. Lakoff’s is an important early contribution that shows some of the actual interaction between embodiment and cognition.

I find it particularly relevant when he talks about how concepts are motivated but not determined by embodied cognition.

Constructions

Lakoff’s work was taking shape alongside Fillmore’s work on construction grammar and Langacker’s on cognitive grammar. While the current construction grammar paradigm is much more influenced by those, I think it is still worth reading Lakoff for his contribution here. Particularly case studies 2 and 3 are great examples of the power of this approach.

Additional chapters of interest

Elaborations of core concepts

Chapters 17 and 18 elaborate on the core concepts in important ways but many people never reach them because they follow a lot of work on philosophical implications.

Chapter 17 on Cognitive Semantics takes another more deeper look at ICMs (idealized cognitive models) across various dimensions.

Chapter 18 deals with the question of how conceptual categories work across languages in the context of relativism. The name of the book is derived from a non-English example but this takes the question of universals and language specificity head on. Perhaps not the in the most comprehensive way (the debate on relativism has moved on) but it illuminates the core concepts further.

Case studies

Case Studies 2 and 3 should be of great interest to linguists. Not because they are perfect but because they show the depth of analysis required of even relatively simple concepts.

Philosophical implications

Lakoff is not shy about placing his work in the context of disrruption of the reigning philosophical paradigm of his (and to a significant extent our) day. Chapter 11 goes into more depth on how he understands the ‘objectivist paradigm’. It has been criticised for not representing actual philosophical positions (which he explicitly says he’s not doing) but I think it’s representative of many actual philosophical and other treatments of language and cognition.

This is then elaborated in chapters 12 – 16 and of course in his subsequent book with Mark Johnson Philosophy in the Flesh. I find the positive argument they’re making compelling but it is let down by staying on the surface of the issues they’re criticising.

What to skip

Where Lakoff (and elsewhere Lakoff and Johnson) most open themselves to criticism is their relatively shallow reading of their opponents. Most philosophers don’t engage with this work because they don’t find it speaks their language and when it does, it is easily dismissed as too light.

While I think that the broad critique this book presents of what it calls ‘objectivist approaches’ is correct, I don’t recommend that anyone takes the details too seriously. Lakoff simultaneously gives it too little and too much attention. He argues against very small details but leaves too many gaps.

This means that those who should be engaging with the very core of the work’s contribution fixate on errors and gaps in his criticism and feel free to dismiss the key aspects of what he has to say (much to their detriment).

For example, his critique of situational semantics leaves too many gaps and left him open to successful rejoinders even if he was probably right.

What is missing

While Lakoff engages with cognitive anthropology (and he and Johnson acknowledge their debts in the preface to Metaphors We Live By), he does not reflect the really interesting work in this area. Goffman (shockingly) gets no mention, nor does Victor Turner whose work on liminality is pretty important companion.

There’s also little acknowledgement of work on texts such as that by Halliday and Hasan (although, that was arguably still waiting for its greatest impact in the mid 1980s with the appearance of corpora). But Lakoff and most of the researchers in this areas stay firmly at the level of a clause. But give that my own work is mostly focusing on discourse and text-level phenomena, I would say that.

What to read next

Here are some suggestions for where to go next for elaborations of the key concepts or ideas with relevance to those outlined in the book.

  • Moral politics by Lakoff launched his forays into political work but I think it’s more important as an example of this way of thinking applied for a real purpose. He replaces Idealized Cognitive Models with Frames but shows many great examples of them at work. Even if it falls short as an exhaustive analysis of the issues, it is very important as a methodological contribution of how frames work in real life. I think of it almost as a fourth case study to this book.
  • The Way We Think by Gilles Fauconnier and Mark Turner provides a model of how cognitive models work ‘online’ during the process of speaking. Although, it has made a more direct impact in the field of construction grammar, its importance is still underappreciated outside it. I think of it as an essential companion to the core contribution of this book. Lakoff himself draws on Fauconnier’s earlier work on mental spaces in this book.
  • Work on construction grammar This book was one of the first places where the notion of ‘construction’ in the sense of ‘construction grammar’ was introduced. It has since developed in its own substantive field of study that has been driven by others. I’d say the work of Adele Goldberg is still the best introduction but for my money William Croft’s ‘Radical Construction Grammar’ is the most important. Taylor’s overview of the related ‘Cognitive Grammar’ is also not a bad next read.
  • Work on cognitive semantics There is much to read here. Talmy’s massive 2 volumes of ‘Cognitive Semantics’ are perhaps the most comprehensive but most of the work here happens across various journals. I’m not aware of a single shorter introduction.
  • Philosophy and the Mirror of Nature by Richard Rorty is a book I frankly wish Lakoff had read. Rorty’s taking apart of philosophy’s epistemological imaginings is very much complementary to Lakoff’s critique of ‘objectivism’ but done while engaging deeply with the philosophical issues. While I basically go along with Lakoff’s and later Lakoff and Johnson’s core argument, I can see why it could be more easily dismissed than Rorty. Of course, Rorty’s work is also better known for its reputation than deeply reflected in much of today’s philosophy. Lakoff and Johnson’s essential misunderstanding of Rorty’s contribution and fundamental compatibility with their project in Philosophy in the Flesh is an example of why so many don’t take that aspect of this work seriously. (Although, they are right that both Rorty and Davidson would have been better served by a less impoverished view of meaning and language.)

What language looks like: Dictionary and grammar are to language what standing on one foot is to running

Background

Sometimes a rather obscure and complex analogy just clicks into place in one’s mind and allows a slightly altered way of thinking that just makes so much sense, it hurts. Like putting glasses on in the morning and the world suddenly snapping into shape.

This happened to me this morning when reading the Notes from Two Scientific Psychologists blog and the post on Do people really not know what running looks like?

It describes the fact that many famous painters (and authors of instructional materials on drawing) did not depict running people correctly. When running, it is natural (and essential) to put forward the arm opposite the leg that’s going forward. But many painters who depict running (including the artist who created the poster for the 1922 Olympics!) do it the wrong way round. Not just the wrong way, the way that is almost impossible to perform. And this has apparently been going for as long depiction has been thing. But it’s not just artists (who could even argue that they have other concerns). What’s more when you ask a modern human being to imitate somebody running in a stationary pose (as somebody did on the website Phoons­) they will almost invariably do it the wrong way round. Why? There are really two separate questions here.

  1. Why don’t the incorrect depictions of running strike most people as odd?
  2. Why don’t we naturally arrange our bodies into the correct stance when asked to imitate running while standing still?

Andrew Wilson (one of the two psychologists) has the perfect answer to question 2:

Asking people to pose as if running is actually asking them to stand on one leg in place, and from their point of view these are two very different things with, potentially, two different solutions. [my emphasis]

And he prefaces that with a crucial point about human behavior:

people’s behaviour is shaped by the demands of the task they are actually solving, and that might not be the task you asked them to do.

Do try this at home, try to imitate a runner standing up, then slowly (mime-like), then speed it up. Standing into the wrong configuration is the natural thing to do. Doing it the ‘right’ way round, is hard. It’s not until I sped up into an actual run that my arms found the opposite motion natural until I could keep track of what was going on any more. I would imagine that this would be the case for most people. In fact, the few pictures I could find of runners arranged standing at the start of the race have most of them also with the ‘wrong’ hand/leg position and they’re not even standing on one leg. (See here and here.)

Which brings us back to the first question. Why does not anybody notice? I personally find it really hard to even identify the wrong static description at a glance. I have to slow down, remember what is correct, then match it to the image. What’s going on. We obviously don’t have any cognitive control over the part of running that controls the movement arms in relation tot he movement of legs. We also don’t have any models or social scripts that pay attention to this sort of thing. It is a matter of conscious effort, a learned behaviour, to recognize these things.

Why is this relevant to language?

If you ask someone to describe a language, they will most likely start telling you about the words and the rules for putting them together. In other words, compiling a dictionary and a grammar. They will say something like: “In Albanian, the word for ‘bread’ is ‘bukë'”. Or they will say something like “English has 1 million words.”, “Czech has no word for training.” or “English has no cases.”

All of these statements reflect a notion of language that has a list of words that looks a little like this:

bread n. = 1. baked good, used for food, 2. metaphor for money, etc.
eat v. = 1. process of ingestion and digestion, 2. metaphor, etc.
people n. plural = human beings

And a grammar that looks a little bit like this.

Sentence = Noun (subj.) + Verb + Noun (obj.)

All of this put together will give us a sentence:

People eat food.

All you need is long enough list of words and enough (but not as many) rules and you got a language.

But as linguists have discovered through not a bit of pain, you don’t have a language. You have something that looks like a language but not something that you can actually speak as a language. It’s very similar to language but it’s not language.

Kind of like the picture of the runner with the arms going in the opposite direction. It looks very much like someone running but it’s not it’s just a picture of them running and the picture is fundamentally wrong. Just not in a way that is at all obvious to most people most of the time.

Why grammars and dictionaries seem like a good portrait of language

So, we can ask the same two questions again.

  1. Why does the stilted representation of language as rules and words not strike most people (incl. Steven Pinker) as odd?
  2. Why don’t we give more realistic examples of language when asked to imitate one?

Let’s start with question 2 again which will also give us a hint as to how to answer question 1.

So why, when asked to give an example of English, am I more likely to give:

John loves Mary.

or

Hello. Thank you. Good bye.

than

Is it cold in here? Could you pass the sugar, please. No no no. I’ll think about it?

It’s because I’m achieving a task that is different from actually speaking the language. When asked to illustrate a language, we’re not communicating anything in the language. So our very posture towards the language changes. We start thinking in equivalencies and left and right sides of the word (word = definition) and building blocks of a sentence. Depending on who we’re speaking to, we’ll choose something very concrete or something immediately useful. We will not think of nuance, speech acts, puns or presupposition.

But the vast majority of our language actions are of the second kind. And many of the examples we give of language are actually good for only one thing: Giving an example of the language. (Such as the famous example from logic ‘A man walks’ which James MacCawley analysed as only being usable in one very remote sense.)

As a result, if we’re given the task of describing language, coming up with something looking like a dictionary and a grammar is the simplest and best way of achieving fullfilling the assignment. If we take a scholarly approach to this task over generations, we end up with something that very much looks like the modern grammars and dictionaries we all know .

The problem is that these don’t really give us “a picture of language”, they give us “a picture of a pose of language” that looks so much like language to our daily perception, that we can’t tell the difference. But in fact, they are exactly the opposite of language looks like.

Now, we’re in much more complex waters than running. Although, I imagine the exact performance of running is in many ways culturally determined, the amount of variation is going to be limited by the very physical nature of the relatively simple task. Language on the other hand, is almost all culture. So, I would expect people in different contexts to give different examples. I read somewhere (can’t track down the reference now) that Indian grammarians tended to give examples of sentences in the imperative. Early Greeks (like Plato) had a much more impoverished view of the sentence than I showed above. And I’m sure there are languages with even more limited metalanguage. However, the general point still stands. The way we tend to think about language is determined by the nature of the task

The key point I’ve repeated over and over (following Michael Hoey) is that grammars and dictionaries are above all texts written in the language. They don’t stand aprt from it. They have their own rules, conventions and inventories of expression. And they are susceptible to the politics and prejudices of their time. Even the OUP. At the same time, they can be very useful tools to developing language skills or dealing with unfamiliar texts. But so does asking a friend or figuring out the meaning in context.

Which brings us to question 1. Why has nobody noticed that language doesn’t quite work that way? The answer is that – just like with running – people have. But only when they try to match the description with something that is right in front of them. Even then, they frequently (and I’m talking about professional linguists like Stephen Pinker here) ignore the discrepancy or ascribe it to a lack of refinement of the descriptions. But most of the time, the tasks that we fulfil with language do not require us to engage the sort of metacognitive aparatus that would direct us to reflect on what’s actually going on.

What does language really look like

So is there a way to have an accurate picture of language? Yes. In fact, we already have it. It’s all of it. We don’t perhaps have all the fine details, but we have enough to see what’s going on – if we look carefully. It’s not like linguists of all stripes have not described pretty much everything that goes on with language in one way or another. The problem is that they try to equate the value of a description to the value of the corresponding model that very often looks like an algorithm amenable to being implemented in a computer program. So, if I describe a phenomenon of language as a linguist, my tendency is to immediately come up with a fancy looking notation that will look like ‘science’. If I can make it ‘mathematical’, all the better. But all of these things are only models. They are ways of achieving a very particular task. Which is to – in one way or another – model language for a particular purpose. Development of AI, writing of pedagogic grammars, compiling word lists, predicting future trends, tracing historical developments, estimating psychological impact, etc. All of these are distinct from actual pure observation of what is going on. Of course, even simple description of what I observe is a task of its own with its own requirements. I have to choose what I notice and decide what I report on. It’s a model of a sort, just like an accurate painting of a runner in motion is just a model (choosing what to emphasize, shadows, background detail, facial expression, etc.) But it’s the task we’re really after: Coming up with as accurate and complete a picture of language as is possible for a collectivity of humans.

People working in construction grammars in the usage-based approach are closest to the task. But they need to talk with people who work on texts, as well, if they really want to start painting a fuller picture.

Language is signs on doors of public restrooms, dirty jokes on TV, mothers speaking to children, politicians making speeches, friends making small talk in the street, newscasters reading the headlines, books sold in bookshops, gestures, teaching ways of communication in the classroom, phone texts, theatre plays, songs, blogs, shopping lists, marketing slogans, etc.

Trying to reduce their portrait to words and rules is just like trying to describe a building by talking about bricks and mortar. They’re necessary and without them nothing would happen. But a building does not look like a collection of bricks and mortar. Nor does knowing how to put a brick to brick and glue them together help in getting a house built. At best, you’d get a knee-high wall. You need a whole of other knowledge and other kinds of strategies of building corners, windows, but also getting a planning permission, digging a foundation, hiring help, etc. All of those are also involved in the edifices we construct with language.

An easy counterargument here would be: That’s all well and good but the job of linguistics is to study the bricks and the mortar and we’ll leave the rest to other disciplines like rhetoric or literature. At least, that’s been Chomsky’s position. But the problem is that even the words and grammar rules don’t actually look like what we think they do. For a start, they’re not arranged in any of the ways in which we’re used to seeing them. But they probably don’t even have the sorts of shapes we think of them in. How do I decide whether I say, “I’m standing in front of the Cathedral” or “The Cathedral is behind me.”? Each of these triggers a very different situation and perspective on exactly the same configuration of reality. And figuring out which is which requires a lot more than just the knowledge of how the sentence is put together. How about novel uses of words that are instantly recognizable like “I sneezed the napkin off the table.” What exactly are all the words and what rules are involved?

Example after example shows us that language does not look very much like that traditional picture we have drawn of it. More and more linguists are looking at language with freshly open eyes but I worry that they may get off task when they’re asked to make a picture what they see.

Where does the metaphor break

Ok, like all metaphors and analogies, even this one must come to an end. The power of a metaphor is not just finding where it fits but also pointing out its limits.

The obvious breaking point here is the level of complexity. Obviously, there’s only one very discretely delineated aspect of what the runners are doing that does not match what’s in the picture. The position of the arms. With language, we’re dealing with many subtle continua.

Also, the notion of the task is taken from a very specific branch of cognitive psychology and it may be inappropriate extending it to areas where tasks take a long time, are collaborative and include a lot of deliberately chosen components as well as automaticity.

But I find it a very powerful metaphor nevertheless. It is not an easy one to explain because both fields are unfamiliar. But I think it’s worth taking the time with it if it opens the eyes of just one more person trying to make a picture of language looks like.

Anthropologists’ metaphorical shenanigans: Or how (not) to research metaphor

Over on the excellent ‘Genealogy of Religion’, Cris Campbell waved a friendly red rag in front of my eyes to make me incensed over exaggerated claims (some) anthropologists make about metaphors. I had expressed some doubts in previous comments but felt that perhaps this particular one deserves its own post.

The book Cris refers to is a collection of essays  America in 1492: The World of the Indian Peoples Before the Arrival of Columbus (1991, ed. Alvin Josephy) which also contains an essay by Joel Sherzer called “A Richness of Voices”.  I don’t have the book but I looked up a few quotes on metaphor from the book.

The introduction summarizes the conclusion thus:

“Metaphors about the relations of people to animals and natural forces were essential to the adaptive strategies of people who lived by hunting.” (p. 26)

This is an example what Sherzer has to say about metaphor:

“Another important feature of native vocabularies was the metaphor – the use of words or groups of words that related to one realm of meaning to another. To students they provide a window into American Indian philosophies. … The relationship between the root and the derived form was often metaphorical.” (p. 256)

The first part of both statements is true but the second part does not follow. That is just bad bad scholarship. I’m not a big Popperian but if you want to make claims about language you have to postulate some hypotheses and try really really really hard to disprove them. Why? Because there are empirical aspects to the questions that can have empirical support. Instead the hypotheses are implied and no attempt is made to see if they work. So this is what I suggest are Sherzer’s implicit hypotheses that should be made explicit and tested:

  1. American Indian languages use metaphors for essential parts of their understanding of the world. (Corollary: If we understand the metaphors, we can understand the worldview of the speakers of those languages.)
  2. American Indian language use of metaphors was necessary to their survival because of their hunter-gatherer lifestyles.
  3. American Indian languages use metaphor than the SAE (Standard Average European) languages.

Re 1: This is demonstrably true. It is true of all languages so it is not surprising here. However, exactly how central this metaphorical reasoning is and how it works cognitively is an open question. I addressed some of this in my review of Verena Haser’s book.

As to the corollary, I’ve mentioned this time and time again. There is no straightforward link between metaphor and worldview. War on poverty, war on drugs, war on terror all draw on different aspects of war. As does Salvation Army, Peace Corps and the Marine Corps. You can’t say that Salvation Army subscribes to the same level of violence than a ‘real’ Army. The same goes for metaphors like ‘modern Crusades’ or the various notions of ‘Jihad’. Metaphor works exactly because it does not commit us to a particular course of action.

That’s not to say that the use of metaphor can never be revealing of underlying conceptualizations. For instance, calling something rebellion vs. calling it a ‘civil war’ imposes a certain order on the configuration of participants and reveals the speaker’s assumptions. But calling one ‘my rock’ does not reveal any cultural preoccupation with rocks. The latter (I propose) is much more common than the former.

Re 2: I think this is demonstrably false. From my (albeit incomplete) reading of the literature, most of the time metaphors just got in the way of hunting. Thinking of the ‘Bear’ as the father to whom you have to ritually apologise before killing him seems a bit excessive. Over metaphorisation of plants and animals has also led to their over or under exploitation. E.g. the Nuer not eating birds and foregoing an important source of nutricion or the Hawains hunting rare birds to extinction for their plumes. Sure, metaphors were essential to the construction of folk taxonomies but that is equally true of Western ‘scientific’ taxonomies which map into notions of descent, progress and containment. (PS: I’ve been working on a post called ‘Taxonomies are metaphors where I elaborate on this).

Re 3: This is just out and out nonsense. The example given are stuff like bark of the tree being called ‘skin’ and spacial prepositions like ‘on top of’ or ‘behind’ being derived from body parts. The author obviously did not bother to consult an English etymological dictionary where he could discover that ‘top’ comes from ‘tuft’ as in ‘tuft of hair’ (or is at the very least connected). And of course the connection of ‘behind’ to body part (albeit in the other direction) should be pretty obvious to anyone. Anyway, body part metaphors are all over all languages in all sorts similar but inconsistent ways: mountains have feet (but not heads), human groups have heads (but not feed), trees have trunks (but not arms), a leader may have someone as their right arm (but not their left foot). And ‘custard has skin’ in English (chew on that). In short, unless the author can show even a hint of a quantitative tendency, it’s clear that American Indian languages are just as metaphorical as any other languages.

Sherzer comes to this conclusion:

“Metaphorical language pervaded the verbal art of the Americas in 1492, in part because of the closeness Native American had always felt to the natural world around them and their social, cultural, aesthetic, and personal identification with it and in part because of their faith in the immediacy of a spirit world whose presence could be manifest in discourse.”

But that displays fundamental misunderstanding of how metaphor works in language. Faith in immediacy has no link to the use of metaphors (or at the very least Sherzer did not demonstrate any link because he confused lyricism with scholarship). Sure, metaphors based on the natural world might indicate ‘closeness to the natural world around’ but that’s just as much of a discovery as saying that people who live in an area with lots of oaks have a word that means ‘oak’. The opposite would be surprising. The problem is that if you analyzed English without preconceptions about the culture of its speakers you would find as much of a closeness to the natural world (e.g a person can be ‘a force of nature’, ‘eyes like a hawk’, ‘dirty as a pig’, ‘wily as a fox’, ‘slow as a snail’, ‘beautiful as a flower’, ‘sturdy as a tree’, etc.).

While this seems deep, it’s actually meaningless.

“The metaphorical and symbolic bent of Mesoamerica was reflected in the grammars, vocabularies, and verbal art of the region. (p. 272)

Mesoamerica had no ‘symbolic bent’. Humans have a symbolic bent, just like they have spleens, guts and little toes.  So let’s stop being all gushy about it and study things that are worth a note.

PS: This just underscores my comments on an earlier post of Cris’ where I took this quote to task:

“Nahuatl was and is a language rich in metaphor, and the Mexica took delight in exploring veiled resemblances…” This is complete and utter nonsense. Language is rich in metaphor and all cultures explore veiled resemblances. That’s just how language works. All I can surmise is that the author did not learn the language very well and therefore was translating some idioms literally. It happens. Or she’s just mindlessly spouting a bullshit trope people trot out when they need to support some mystical theory about a people.

And the conclusion!? “In a differently conceptualized world concepts are differently distributed. If we want to know the metaphors our subjects lived by, we need first to know how the language scanned actuality. Linguistic messages in foreign (or in familiar) tongues require not only decoding, but interpretation.” Translated from bullshit to normal speak: “When you translate things from a foreign language, you need to pay attention to context.” Nahuatl is no different to Spanish in this. In fact, the same applies to British and American English.

Finally, this metaphor mania is not unique to anthropologists. I’ve seen this in philosophy, education studies, etc. Metaphors are seductive… Can’t live without them…

Image by moune.drah CC BY NC SA

What does it mean when words ‘really’ mean something: Dismiss the Miss

A few days ago, I tweeted a link to an article in TES:

Today, I got the following response back:

@lizzielh is absolutely right. As the title of an as yet unpublished blog post of mine goes: “Words don’t mean things, people mean things”. I even wrote a whole book chapter on that with the same title as this post.

Indeed, if it had been me writing on the topic, I would have chosen a more judicious title. Such as “The legacy of discrimination behind the humble Miss” or “Past and present inequalities encoded in the simple Miss”.

In fact, the only reason I tweeted that article in the first place was because it was making a much more subtle and powerful point than simple etymology (as you would expect from one based on the work of the eminent scholar of language and gender Jennifer Coates). Going all the way back to Language and the Woman’s Place and even before, people have been trying to peg the blame on simple words. All along the response has been, but these are just words, we don’t mean anything bad by them. Or, these are just words, the real harm is done in the real world.

Many women I meet continue to like the Miss/Mrs distinction despite the long availability of the now destigmatized Ms. It was not too long ago that I set up a sign up form with only Prof Dr Mr Ms and got lots of complaints from women who wanted to keep their Miss or Mrs. So restigmatizing Miss is actively harmful to the self-image of many women whose identity is tied with that label. Feminist tend to make light of the ‘unfeminist’ cry of “I like it when men open the door to me”, or “Carrying my bag for me just shows respect”.  Or going back even further, “I don’t need a vote, I exercise my influence through my husband.” But change is literally hard, it takes time and effort, so an attempt at making the world better will always making temporarily worse (at least for some people).

The fact is that Miss is a bound in a network of meanings, interactions and power relations. And even if it takes some mental pain, it’s worth picking at all it covers up.

But not every minute of every day. Sometimes, we need to say something to get from conversational point A to conversational point B and even a laden word may be better than no word. As one of the respondents in the article says:

My response is always that my name isn’t Miss; it’s Mrs Coslett. But if I’m in a school where students don’t know me and they call me Miss, I’m fine with that. They’re showing respect by giving me a title, rather than ‘hey’ or ‘oi, you’ or whatever.

Most of the time contentious words are used, challenging them is not feasible. But she’s wrong in her conclusion:

That’s just the way the English language works.

That’s absolutely not true. Just like words don’t mean anything on their own, language does not just work. It’s used to do things (to riff on Austin’s famous book) by people. It is not always used purposefully but its use is always bound in the many ways and means of people. The way we speak now is a result of centuries of little power plays, imitations of prestige, prescriptions of obedience. When you look closer, they’re all easy to see.

Things have let up considerably since the 1970s. Many fewer people are concerned about how language encodes gender inequality and it’s worthwhile reminding ourselves that many of the historical unfairnesses hidden in word histories are still with us. Just like you can’t get away with saying “I didn’t mean anything by the ‘n’ word”, you can’t just shrug off the critique of the complex tapestry of gender bias in ‘Miss’.

Miss does not “really mean” anything. It’s just a sequence of letters or sounds. And most people using it do not “really mean” anything by it. Or it does not “really mean” anything to them. But context is everything.

It is a truism to say that racism will be done away with when people don’t dislike each other because of the color of their skin. But the opposite is the case. The sign that racism has disappeared is when I can say “I really don’t like black people” simply because I don’t like the color of their skin in the same way I may prefer redheads to blondes. Preference for skin colour is then just a harmless quirk. But we’re centuries away from that because any such preference is tied to a system of discrimination going back a long way.  (BTW: just to avoid misunderstanding, I personally find black skin beautiful.)

The same thing applies to “Miss”, we can’t just turn our back on its pernicious potential. Most of the time it’s hidden from sight but it’s recoverable at a moment’s notice. Because we live in a world where male is still the default position. We have to work to change that. Change our minds, hearts, cognitions and languages. They don’t  just work on their own. We make them work. So let’s make them work for us. The ‘us’ we want to be, rather than the ‘us’ we used to be in the bad old days.

Photo Credit: abdallahh via Compfight cc

What is not a metaphor: Modelling the world through language, thought, science, or action

The role of metaphor in science debate (Background)

Recently, the LSE podcast an interesting panel on the subject of “Metaphors and Science”. It featured three speakers talking about the interface between metaphor and various ‘scientific’ disciplines (economics, physics and surgery). Unlike many such occasions, all speakers were actually very knowledgeable and thoughtful on the subject.

In particular, I liked Felicity Mellor and Richard Bronk who adopted the same perspective that underlies this blog and which I most recently articulated in writing about obliging metaphors. Felicity Mellor put it especially eloquently when she said:

“Metaphor allows us to speak the truth by saying something that is wrong. That means it can be creatively liberating but it can also be surreptitiously coercive.”

This dual nature of coerciveness and liberation was echoed throughout the discussion by all three speakers. But they also shared the view of ubiquity of metaphor which is what this post is about.

What is not a metaphor? The question!

The moderator of the discussion was much more stereotypically ambivalent about such expansive attitude toward metaphor and challenged the speakers with the question of ‘what is the opposite of metaphor’ or ‘what is not a metaphor’. He elicited suggestions from the audience, who came up with this list:

model, diagram, definition, truths, math, experience, facts, logic, the object, denotation

The interesting thing is that most of the items on this list are in fact metaphorical in nature. Most certainly models, diagrams and definitions (more on these in future posts). But mathematics and logic are also deeply metaphorical (both in their application but also internally; e.g. the whole logico mathematical concept of proof is profoundly metaphorical).

Things get a bit more problematic with things like truth, fact, denotation and the object. All of those seem to be pointing at something that is intuitively unmetaphorical. But it doesn’t take a lot of effort to see that ‘something metaphorical’ is going on there as well. When we assign a label (denotation), for instance, ‘chair’ or ‘coast’ or ‘truth’ we automatically trigger an entire cognitive armoury for dealing with things that exist and these things have certain properties. But it is clear that ‘chair’, ‘coast’ and ‘metaphor’ are not the same kind of thing at all. Yet, we can start treating them the same way because they are both labels. So we start asking for the location, shape or definition of metaphor, just because we assigned it a label in the same way we can ask for the same thing about a chair or a coast. We want to take a measure of it, but this is much easier with a chair than with a coast (thus the famous fractal puzzle about the length of the coast of Britain). But chairs are not particularly easy to nail down (metaphorically, of course) either, as I discussed in my post on clichés and metaphors.

Brute facts of tiny ontology

So what is the thing that is not a metaphor? Both Bronk and Mellor suggested the “brute fact”. A position George Lakoff called basic realism and I’ve recently come to think of as tiny ontology. The idea, as expressed by Mellor and Bronk in this discussion, is that there’s a real world out there which impinges upon our bodily existence but with which we can only interact through the lens of our cognition which is profoundly metaphorical.

But ultimately, this does not give us a very useful answer. Either everything is a metaphor, so we might as well stop talking about it, or there is something that is not a metaphor. In which case, let’s have a look. Tiny ontology does not give us the solution because we can only access it through the filter of our cognition (which does not mean consciously or through some wilful interpretation). So the real question is, are there some aspects of our cognition that are not metaphorical?

Metaphor as model (or What is metaphor)

The solution lies in the revelation hinted at above that labels are in themselves metaphors. The act of labelling is metaphorical, or rather, it triggers the domain of objects. What do I mean by that? Well, first let’s have a look at how metaphor actually works. I find it interesting that nobody during the entire discussion tried to raise that question other than the usual ‘using something to talk about something else’. Here’s my potted summary of how metaphor works (see more details in the About section).

Metaphor is a process of projecting one conceptual domain onto another. All of our cognition involves this process of conceptual integration (or blending). This integration is fluid, fuzzy and partial. In language, this domain mapping is revealed through the process of deixis, attribution, predication, definition, comparison, etc. Sometimes it is made explicit by figurative language. Figurative language spans the scale of overt to covert. Figurative language has a conceptual, communicative and textual dimension (see my classification of metaphor use). In cognition, this process of conceptual integration is involved in identification, discrimination, manipulation. All of these can be more or less overtly analogical.

So all of this is just a long way of saying, that metaphor is a metaphor for a complicated process which is largely unconscious but not uncommonly conscious. In fact, in my research, I no longer use the term ‘metaphor’ because it misleads more than it helps. There’s simply too much baggage from what is just overt textual manifestation of metaphor – the sort of ‘common sense’ understanding of metaphor. However, this common sense ordinary understanding of ‘metaphor’ makes using the word a useful shortcut in communication with people who don’t have much of a background in this thought. But when we think about the issue more deeply, it becomes a hindrance because of all the different types of uses of metaphor I described here (a replay of the dual liberating and coercive nature of metaphor mentioned above – we don’t get escape our cognition just because we’re talking about metaphors).

In my work, I use the term frame, which is just a label for a sort of conceptual model (originally suggested by Lakoff as Idealized Cognitive Model). But I think in this context the term ‘model’ is a bit more revealing about what is going on.

So we can say that every time, we engage conceptually with our experience, we are engaging in an act of modelling (or framing). Even when I call something ‘true’, I am applying a certain model (frame) that will engage certain heuristics (e.g. asking for confirmation, evidence). Equally, if I say something like ‘education is business’, I am applying a certain model that will allow me to talk about things like achieving economies of scale or meeting consumer demand but will make it much harder to talk about ethics and personal growth. That doesn’t mean that I cannot apply more than one model, a combination of models or build new models from old ones. (Computer virus is a famous example, but natural law is another one. Again more on this in later posts.)

Action as an example of modelling

The question was asked during the discussion by an audience member, whether I can experience the world directly (not mediated by metaphoric cognition). The answer is yes, but even this kind of experience involves modelling. When I walk along a path, I automatically turn to avoid objects – therefore I’m modelling their solid and interactive nature. Even when I’m lying still, free of all thought and just letting the warmth of the shining sun wash over me, I’m still applying a model of my position in the world in a particular way. That is, my body is not activating my ears to hear the sun rays, nor is it perceiving the bacteria going about their business in my stomach. A snake, polar bear or a fish would all model that situation in a different way.

This may seem like unnecessary extension of the notion of a model. (But it echos the position of the third speaker Roger Kneebone who was talking about metaphor as part of the practice of surgery.) It is not particularly crucial to our understanding of metaphor, but I think it’s important to divert us from a certain kind of perceptual mysticism in which many people unhappy with the limitations of their cognitive models engage. The point is that not all of our existence is necessarily conceptual but all of it models our interaction with the world and switches between different models as appropriate. E.g. my body applies different models of the world when I’m stepping down from a step on solid ground or stepping into a pool of water.

The languages of metaphor: Or how a metaphor do

I am aware that this is all very dense and requires a lot more elaboration (well, that’s why I’m writing a blog, after all). But I’d like to conclude with a warning that the language used for talking about metaphor brings with it certain models of thinking about the world which can be very confusing if we don’t let go of them in time. Just the fact that we’re using words is a problem. When words are isolated (for instance, in a dictionary or at the end of the phrase ‘What is a…’) it only seems natural that they should have a definition. We have a word “metaphor” and it would seem that it needs to have some clear meaning. The kind of thing we’re used to seeing on the right-hand side of dictionaries. But insisting that dictionary-like definition is what must be at the end of the journey is to misunderstand what we’ve seen along the way.

There are many contexts in which the construction “metaphor is…” is not only helpful but also necessary. For example when clarifying one’s use: “In this blog, what I mean by metaphor is much broader than what traditional students of figurative language meant by it.” But in the context of trying to get at what’s going on in the space that we intuitively describe as metaphorical, we should almost be looking for something along to the lines of “metaphor does” or “metaphors feels like”. Or perhaps refrain from the construction “metaphor verb” altogether and just admit that we’re operating in a kind of metaphor-tasting soup. We can get at the meaning/definition by disciplined exploration and conversation.

In conclusion, metaphor is a very useful model when thinking about cognition, but it soon fails us, so we can replace it with more complex models, like that of a model. We are then left with the rather unsatisfactory notion of a metaphor of metaphor or a model of model. The sort of dissatisfaction that lead Derrida and his like to the heights of obscurity. I think we can probably just about avoid deconstructionist obscurantism but only if we adopt one of its most powerful tools, the fleeting sidelong glance (itself a metaphor/model). Just like the Necker cube, this life on the edge of metaphor is constantly shifting before our eyes. Never quite available to us perceptually all at once but readily apprehended by us in its totality. At once opaque and so so crystal clear. Rejoice all you parents of freshly screaming thoughts. It’s a metaphor!
Photo Credit: @Doug88888 via Compfight cc

Linguistics according to Fillmore

While people keep banging on about Chomsky as being the be all and end all of linguistics (I’m looking at you philosophers of language), there have been many linguists who have had a much more substantial impact on how we actually think about language in a way that matters. In my post on why Chomsky is not really a linguist at all I listed a few.

Sadly, one of these linguists died yesterday. It was Charles J Fillmore who was a towering figure among linguists without writing a single book. In my mind, he changed the face of linguistics three times with just three articles (one of them co-authored). Obviously, he wrote many more but compared to his massive impact, his output was relatively modest. His ideas have been with me all through my life as a linguist and on reflection, they form a foundation about what I know language to be. Therefore, this is not so much an obituary (for which I’m hardly the most qualified person out there) as a manifesto for a linguistics of a truly human language.

The case for Fillmore

The first article, more of a slim monograph at 80 odd pages, was Case for Case (which, for some reason, I first read in Russian translation). Published in 1968 it was one of the first efforts to find deeper functional connections in generative grammar (following on his earlier work with transformations). If you’ve studied Chomskean Government and Binding, this is where thematic roles essentially come from. I only started studying linguistics in 1991 which is when Case for Case was already considered a classic. Particularly in Prague where function was so important. But even after all those years, it is still worth reading for any minimalist  out there. Unlike so many in today’s divided world, Fillmore engaged with the whole universe of linguistics, citing Halliday, Tesniere, Jakobson,  Whorf, Jespersen, and others while giving an excellent overview of the treatment of case by different theories and theorists. But the engagement went even deeper, the whole notion of ‘case’ as one “base component of the grammar of every language” brought so much traditional grammar back into contact with a linguistics that was speeding away from all that came before at a rate of knots.

From today’s perspective, its emphasis on the deep and surface structures, as well as its relatively impoverished semantics may seem a bit dated, but it represents an engagement with language used to express real meaning.  The thinking that went into deep cases transformed into what has become known as Frame Semantics (“I thought of each case frame as characterizing a small abstract ‘scene’ or ‘situation’, so that to understand the semantic structure of the verb it was necessary to understand the properties of such schematized scenes” [1982]) which is where things really get interesting.

Fillmore in the frame

When I think about frame semantics, I always go to his 1982 article Frame Semantics published in the charmingly named conference proceedings ‘Linguistics in the morning calm’ but it had its first outing in 1976. George Lakoff used it as one of the key inspirations to his idealized cognitive models in Women, Fire, and Dangerous things which is where this site can trace its roots. As I have said before, I essentially think about metaphors as a special kinds of frames.

In it, he says:

By the term ‘frame’ I have in mind any system of concepts related in such a way that to understand anyone of them you have to  understand the whole structure in which it fits; when one of the things in such a structure is introduced into a text, or into a conversation, all of the others are automatically made available. I intend the word ‘frame’ as used here to be a general cover term for the set of concepts variously known, in the literature on natural language understanding, as ‘schema: ‘script’, ‘scenario’, ‘ideational scaffolding’, ‘cognitive model’, or ‘folk theory’.

It is a bit of a mouthful but it captures in a paragraph the absolute fundamentals of the semantics of human language as opposed to projecting the rules of formal logic and truth conditions onto an impoverished version of language that all the generative-inspired approaches try to do. Also, it brings together many other concepts from different fields of scholarship. Last year I presented a paper on the power of the concept of frame where I found even more terms that have a close affinity to it which only underscores the far reaching consequences of Fillmore’s insight.

As I was looking for some more quotes from that article, I realized that I’d have to pretty much cut and paste in the whole of it. Almost, every sentence there is pure gold. Rereading it now after many many years, it’s becoming clear how many things from it I’ve internalized (and frankly, reinvented some of the ideas I forgot had been there).

Constructing Fillmore

About the same time, and merging the two earlier insights, Fillmore started working on the principles that have come to be known as construction grammar. Although, by then, the ideas were some years old, I always think of his 1988 article with Paul Kay and Mary Catherine O’Conner as a proper construction grammar manifesto. In it they say:

The overarching claim is that the proper units of a grammar are more similar to the notion of construction in traditional and pedagogical grammars than to that of rule in most versions of generative grammar.

Constructions, according to Fillmore have these properties:

  1. They are not limited to the constituents of a single syntactic tree. Meaning, they span what has been considered as the building blocks of language.
  2. They specify at the same time syntactic, lexical, semantic and pragmatic information.

  3. Lexical items can also be viewed as constructions (this is absolutely earth shattering and I don’t think linguistics has come to grips with it, yet).

  4. They are idiomatic. That is, their meaning is not built up from their constituent parts.

Although Lakoff’s study of ‘there constructions’ in Women, Fire, and Dangerous Things came out a year earlier (and is still essential reading), I prefer Fillmore as an introduction to the subject (if only because I never had to translate it).

The beauty of construction grammar (just as the beauty of frame semantics) is in that it can bridge much of the modern thinking about language with grammatical insights and intuitions of generations of researchers from across many schools of thought. But I am genuinely inspired by its commitment to language as a whole, expressed in the 1999 article by Fillmore and Kay:

To adopt a constructional approach is to undertake a commitment in principle to account for the entirety of each language. This means that the relatively general patterns of the language, such as the one licensing the ordering of a finite auxiliary verb before its subject in English as illustrated in 1, and the more idiomatic patterns, such as those exemplified in 2, stand on an equal footing as data for which the grammar  must provide an account.

(1) a. What have you done?  b. Never will I leave you. c. So will she. d. Long may you prosper! e. Had I known, . . . f. Am I tired! g. . . . as were the others h. Thus did the hen reward Beecher.

(2) a. by and large b. [to] have a field day c. [to] have to hand it to [someone]  d. (*A/*The) Fool that I was, . . . e. in x’s own right

Given such a commitment, the construction grammarian is required to develop an explicit system of representation, capable of encoding economically and without loss of generalization all the constructions (or patterns) of the language, from the most idiomatic to the most general.

Notice that they don’t just say ‘language’ but ‘each language’. Both of those articles give ample examples of how constructions work and what they do and I commend them to your linguistic enjoyment.

Ultimately, I do not subscribe to the exact version of construction grammar that Fillmore and Kay propose, agreeing with William Croft that it is still too beholden to the formalist tradition of the generative era, but there is something to learn from on every page of everything Fillmore wrote.

Once more with meaning: the FrameNet years

Both frame semantics and construction grammar impacted Fillmore’s work in lexicography with Sue Atkins and culminated in FrameNet a machine readable frame semantic dictionary providing a model for a semantic module to a construction grammar. To make the story complete, we can even see FrameNet as a culmination of the research project begun in Case for Case  which was the development of a “valence dictionary” (as he summarized it in 1982). While FrameNet is much more than that and has very much abandoned the claim to universal deep structures, it can be seen as accomplishing the mission of a language with meaning Fillmore set out on in the 1960s.

Remembering Fillmore

I only met Fillmore once when he came to lecture at a summer school in Prague almost twenty years ago. I enjoyed his lectures but was really too star struck to take advantage of the opportunity. But I saw enough of him to understand why he is remembered with deep affection and admiration by all of his colleagues and students whose ranks form a veritable who’s who of linguists to pay attention to.

In my earlier post, I compared him in stature and importance to Roman Jakobson (even if Jakobson’s crazy voluminous output across four languages dwarfs Fillmore’s – and almost everyone else’s). Fillmore was more than a linguist’s linguist, he was a linguist who mattered (and matters) to anyone who wanted (and wants) to understand how language works beyond a few minimalist soundbites. Sadly it is possible to meet graduates with linguistics degrees who never heard of Jakobson or Fillmore. While it’s almost impossible to meet someone who doesn’t know anything about language but has heard of Chomsky. But I have no doubt that in the decades of language scholarship to come, it will be Fillmore and his ideas that will be the foundation upon which the edifice of linguistics will rest. May he rest in peace.

Post Script

I am far from being an expert on Fillmore’s work and life. This post reflects my personal perspective and lessons I’ve learned rather than a comprehensive or objective reference work. I may have been rather free with the narrative arc of his work. Please be free with corrections and clarifications. Language Log reposted a more complete profile of his life.

References

  • Fillmore, C., 1968. The Case for Case. In E. Bach & R. Harms, eds. Universals in Linguistic Theory. New York: Holt, Rinehart and Winston, pp. 1–88. Available at: http://pdf.thepdfportal.com/PDFFiles/123480.pdf [Accessed February 15, 2014].
  • Fillmore, C.J., 1976. Frame Semantics and the nature of language. Annals of the New York Academy of Sciences, 280 (Origins and Evolution of Language and Speech), pp.20–32.
  • Fillmore, C., 1982. Frame Semantics. In The Linguistic Society of Korea, ed. Linguistics in the morning calm : International conference on linguistics : Selected papers. Seoul  Korea: Hanshin Pub. Co., pp. 111–139.
  • Fillmore, C.J., Kay, P. & O’Connor, M.C., 1988. Regularity and Idiomaticity in Grammatical Constructions: The Case of Let Alone. Language, 64(3), pp.501–538.
  • Kay, P. & Fillmore, C.J., 1999. Grammatical constructions and linguistic generalizations: the What’s X doing Y? construction. Language, 75(1), pp.1–33.

5 things everybody should know about language: Outline of linguistics’ contribution to the liberal arts curriculum

Drafty

This was written in some haste and needs further refinement. Maybe one day that will come. For now, it will be left as it stands.

Background

This post outlines what I think are the key learnings from the output of the research of the field of linguistics that should be reflected in the general curriculum (in as much as any should be). This is in reaction to the recent posts by Mark Liberman suggesting the role and form of grammar analysis in general education. I argue that he is almost entirely wrong in his assumptions as well as in his emphasis. I will outline my arguments against his position at the end of the post. At the beginning I will outline key easily digestible lessons of modern linguistics that should be incorporated into language education at all levels.

I should note that despite my vociferous disagreement, Mark Liberman is one of my heros. His ‘Breakfast Experiments(tm)’ have brought me much joy and his and his fellow contributors to the Language Log make me better informed about developments in linguistics outside my own specialty that I would otherwise not know about. Thanks Mark for all your great work.

I have addressed some of these issues in previous posts here, here and here.

What should linguistics teach us

In my post on what proponents of simple language should know about linguistics, I made a list of findings that I propose are far more important than specific grammatical and lexicographic knowledge. Here I take a slightly more high-level approach – but in part, this is a repetition of that post.

Simply, I propose that any school-level curriculum of language education should 1. expose students (starting at primary level) to the following 5 principles through reflection on relevant examples, and 2. these principles should be reflected in the practical instruction students receive toward the acquisition of skills and general facility in the standards of that language.

Summary of key principles

  1. Language is a dialect with an army and a navy
  2. Standard English is just one of the many dialects of English
  3. We are all multilingual in many different ways
  4. A dictionary is just another text written in the language, not a law of the language
  5. Language is more than words and rules

Principle 1: Language is a dialect with an army and a navy

This famous dictum (see Wikipedia on origins ) encapsulates the fact that language does not have clear boundaries and that there is no formula for distinguishing where one language ends and another begins. Often, this disctinction depends on the political interests of different groups. In different political contexts, the different Englishes around the world today, could easily qualify for separate language status and many of them have achieved this.

But exploring the examples that help us make sense of this pithy phrase also teaches us the importance of language in the negotiation of our identity and its embeddedness in the wider social sphere. There are piles and piles of evidence to support this claim and learning about the evidence has the potential of making us all better human beings less prone to disenfranchise others based on the way they speak (in as much any form of schooling is capable of such a thing). Certainly more worthy than knowing how to tell the passive voice.

Principle 2: Standard English is just one of the many dialects of English

Not only are there no clear boundaries between languages, there are no clear principles of what constitutes an individual language. A language is identified by its context of use as much as by the forms it uses. So if kayak and a propos can be a part of English so can ain’t and he don’t. It is only a combination of subconscious convention and conscious politics that decides which is which.

Anybody exploring the truth of this statement (and, yes, I’m perfectly willing to say the word truth in this context) will learn about the many features of English and all human languages such as:

  • stratification of language through registers
  • regional and social variation in language
  • processes of change in language over time
  • what we call good grammar are more or less fixed conventions of expression in certain contexts
  • the ubiquity of multiple codes and constant switching between codes (in fact, I think this is so important that it gets a top billing in this list as number 3)

Again, althoguh I’m highly skeptical of claims to causality from education to social change, I can’t see why instruction in this reality of our lives could not contribute to an international conversation about language politics. Perhaps, an awareness of this ‘mantra’ could reduce the frequency of statements such as:

  • I know I don’t speak very good English
  • Word/expression X is bad English
  • Non-native speaker X speaks better English than native speaker Y

And just maybe, teachers of English will stop abusing their students with ‘this is bad grammar’ and instead guide them towards understanding that in different contexts, different uses are appropriate. Even at the most elementary levels, children can have fun learning to speak like a newscaster or a local farm hand, without the violent intrusion into their identity that comes from the misguided and evil labeling of the first as proper and the second as ‘not good English’. Or how about giving the general public enough information to have judged the abominable behavior of the the journalist pseudo elites during the ‘Ebonics controversy’ as the disgraceful display of shameful ignorance it was.

Only and only when they have learned all that, should we mention something about the direct object.

Principle 3: We are all multilingual in many different ways

One of the things linguistics has gathered huge amounts of evidence about is the fact that we are all constantly dealing with multiple quite distinct codes. This is generally not expressed in quite as stark terms as I do here, but I take my cue from bilingualism studies where it has been suggested (either by Chaika or Romaine – I can’t track down the reference to save my life) that we should treat all our study of language as if bilingualism was the default state rather than some exception. This would make good sense even if we went by the educated guess that just over half of the world’s population speaks regularly two or more languages. But I want to go further.

First, we know from principle 1 that there is no definition of language that allows us draw clear boundaries between individual languages. Second, we know from principle 2 that each language consists of many different ‘sub-languages’ or ‘codes’. Because language is so vast and complex, it follows that knowing a language is not an either/or proposition. People are constantly straddling boundaries between different ways of speaking and understanding. Speaking in different ways for different purposes, to different people in different codes. And we know that people switch between the codes constantly for different reasons, even in the same sentence or just one word (very common in languages with rich morphologies like Czech – less common in English but possible with ‘un-fucking-convinving’). Some examples that should illustrate this: “Ladies and gentlemen, we’re screwed” and “And then Jeff said unto Karen”

We also know from all the wailing and gnashing of teeth derriving from the ignorance of principle 2, that acquiring these different codes is not easy. The linguist Jim Miller has suggested to me that children entering school are in a way required to learn a foreign language. In Czech schools, they are instructed in a new lexicon and new morphology (e.g. say ‘malý’ instead of ‘malej’). in English schools they are taught a strange syntax with among other things a focus on nominal structures (cf. ‘he went and’ vs. ‘his going was’) as well as an alien lexicon (cf. ‘leave’ vs. ‘depart’). This is compounded with a spelling system whose principles are often explained on the basis of a phonology they don’t understand (e.g. much of England pronuncing ‘bus’ and ‘booss’ but using teaching materials that rhyme ‘bus’ with ‘us’).

It is not therefore a huge leap to say that for all intents and purposes, we are all multilingual even if we only officially speak one language with its own army and a navy. Or at least, we enagage all the social, cognitive and linguistic processes that are involved in speaking multiple languages. (There is some counter evidence from brain imaging but in my view it is still too early to interpret this either way.)

But no matter whether we accept the strong or the weak version of my proposition, learning about the different pros and cons would make students’ lives much easier at all levels. Instead of feeling like failures over their grammar, they could be encouraged to practice switching between codes. They could also take comfort in the knowledge that there are many different ways of knowing a language and no one person can know it all.

If any time is left over, let’s have a look at constituent structures.

Principle 4: A dictionary is just another text written in the language, not a law of the language

The defference shown to ‘official’ reference materials is at the heart of a myth that the presense of a word in a dictionary in some way validates the word as being a ‘real’ word in the language. But the absolute truth about language that everyone should know and repeat as a mantra every time they ask ‘is X a word’ is that dictionaries are just another text. In fact, they are their own genre of a type that Michael Hoey calls text colonies. This makes them cousins of the venerable shopping list. Dictionaries have their own conventions, their own syntax and their own lexicon. They have ‘heads’ and ‘definitions’ that are both presented in particular ways.

What they most emphatically do not do is confirm or disconfirm the existence of a word or its meaning. It’s not just that they are always behind current usage, it’s that they only reflect a fraction of the knowledge involved in knowing and using words (or as the philosopher John Austin would say ‘doing things with words’). Dictionaries fullfil two roles at once. They are useful tools for gathering information to enable us to deal with the consequences of principle 3 (i.e. to function in a complex multi-codal linguistic environment both as passive and active participants). And they help us express many beliefs about our world such as:

  • The world is composed of entities with meanings
  • Our knowledge is composed of discrete items
  • Some things are proper and others are improper

Perhaps this can become more transparent when we look at entries for words like ‘the’ or ‘cat’. No dictionary definition can help us with ‘the’ unless we can already use it. In this case, the dictionary serves no useful role other than as a catalog of our reality. Performatively assuring us of its own relevance by its irrelevance. How about the entry for ‘cat’. Here, the dictionary can play a very useful role in a bilingual situation. A German will see ‘cat = Katze’ and all will be clear in an instant. A picture can be helpful to those who have no language yet (little children). But the definition of ‘cat’ as “a small domesticated carnivorous mammal with soft fur, a short snout, and retractile claws” is of no use to anybody who doesn’t already know what ‘cat’ means. Or at the very least, if you don’t know ‘cat’, your chances of understanding any definition in the dictionary are very low. A dictionary can be helpful in reminding us that ‘cat’ is also used to refer to ‘man’ among jazz musicians (as in “he’s a cool cat”) but again, all that requires existing knowledge of cat. A dictionary definition that would say ‘a cat is that thing you know as a cat but jazz musicians sometimes use cat to refer to men’ would be just as useful.

In this way, a dictionary is like an audience in the theatre, who are simultaneously watching a performance, and performing themselves the roles of theatre audiences (dress, behavior, speech).

It is also worthwhile to think about what is required of the dictionary author. While the basic part of the lexicographer’s craft is the collection of usage examples (on index cards in the past and in corpora today) and their interpretation, all this requires a prior facility with the language and much introspection about the dictionary makers own linguistic intuitions. So lexicographers make mistakes. Furthermore, in the last hundred years or so, they also almost never start from scratch. Most dictionaries are based on some older dictionaries and the order of definitions is often as much a reflection of a tradition (e.g. in the case of the word ‘literally’ or the word ‘brandish’) as analysis of actual usage.

Why should this be taught as part of the language education curriculum? Simple! Educated people should know how the basic tools surrounding their daily lives work. But even more importantly, they should never use the presence of a word in a dictionary, and as importantly the definition of a word in a dictionary, as the definitive argument for their preferred meaning of a word. (Outside some contexts such as playing SCRABBLE or confirming an uncertainty over archaic or specialist words).

An educated person should be able to go and confirm any guidance found in a dictionary by searching a corpus and evaluate the evidence. It’s not nearly as hard as as identifying parts of speech in a sentence and about a million times more useful for the individual and beneficial for society.

Principle 5: Language is more than words and rules

Steven Pinker immortalised the traditional structuralist vision of what language consists of in the title of his book “Words and rules”. This vision is almost certainly wrong. It is based on an old articulation of language as being the product of a relatively small number of rules applied to a really large number of words (Chomsky expressed this quite starkly but the roots of this model go much deeper).

That is not to say that words and rules are not useful heuristic shortcuts to talking about language. I use this metaphor myself every day. And I certainly am not proposing that language should not be taught with reference to this metaphor.

However, this is a very impoverished view of language and rather than spend time on learning the ‘relatively few’ rules for no good reason other than to please Mark Liberman, why not learn some facts we know about the vastness and complexity of language. That way instead of having a completely misguided view of language as something finite that can be captured in a few simple precepts (often expressed in one of those moronic ‘Top X grammatical errors lists’), one could actually have a basic understanding of all the ways language expresses our minds and impresses itself on our life. Perhaps, we could even get to a generation of psycholinguists and NLP specialists who try to deal with language as it actually is rather than in its bastardised form that can be captured by rules and words.

Ok, so I’m hoisting my theoretical flag here, flying the colors of the ‘usage-based’, ‘construction grammar’, ‘cognitive semantics’ crowd. But the actual curricular proposal is theory free (other than in the ‘ought’ portion of it) and based on well-known and oft-described facts – many of them by the Language Log itself.

To illustrate the argument, let’s open the dictionary and have a look at the entry ‘get’. It will go on for several pages even if we decide to hide all its phrasal friends under separate entries. Wiktionary lists 26 definitions as a verb and 4 as a noun which is fairly conservative. But each of these definitions also comes with usage examples and usage exceptions. For instance, in ‘get behind him’, it is intransitive but in ‘get Jim to come’, it is transitive. This is combined with general rules that apply across all uses such ‘got’ as the past tense and ‘gets’ as the third person singular. Things can be even more complex as with the word ‘bad’ which has the irregular superlative ‘worst’ when it is used in a negative sense as in ‘teaching grammar in schools is the worst idea’ and ‘baddest’ in the positive sense as in ‘Mark Liberman is the baddest linguist on the internet’. ‘Baddest’ is also only appropriate in certain contexts (so my example is at the same time an illustration of code mixing). When we look at any single word in the dictionary, the amount of conscious and unconscious knowledge required to use the word in our daily speech is staggering. This is made even trickier by the fact that not everyone in any one speech community has exactly the same grasp of the word leading to a lot of negotiation and conversation repair.

It is also the sort of stuff that makes understanding of novel expressions like ‘she sneezed the napking off the table’ possible. If we must, let’s do some sentence diagramming now.

Some other things to know

I could go on, some of my other candidate principles that didn’t make this list either because they could be subsumed by one of the items, or they are too theory laden, or because I wanter a list of 5, or because this blog post is over 3,000 words already, are:

  • All lexical knowledge is encyclopedic knowledge
  • Rules of the road like conversation repair, turn taking or text cohesion are just as much part of language as things like passives, etc.
  • Metaphors (and other types of figurative language) are normal, ubiquitous and necessary for language
  • Pretty much every prejudice about gender and language is wrong (like who is more conservative, etc.)
  • No language is more beatiful or amazing than any other, saying this is most likely part of a nationalistic discourse
  • Children are not very good language learners when you put them in the same learning context as adults (e.g. two hours of instruction a week as opposed to living in a culture with no other obligation but to learn)
  • Learning a language is hard and it takes time
  • The etymology of a word does not reflect some deeper meaning of the word
  • Outside some very specific contexts (e.g. language death), languages don’t decline, they change
  • Etc.

Why we should not teach grammar in schools

So, that was my outline of what linguistic expertise should be part of the language education curriculum – and as importantly should inform teachers across all subjects. Now, let’s have a look, as promised, at why Mark Liberman is wrong to call for the teaching of grammar in schools in the first place.

To his credit, he does not trot out any of the usual utilitarian arguments for the teaching of grammar:

  • It will make learning of foreign languages easier
  • It will make today’s graduates better able to express themselves
  • It will contribute to higher quality of discourse
  • It will stop the decline of English
  • It will improve critical thinking of all students

These are all bogus, not supported by evidence and with some evidence against them (see this report for a summary of a part of them).

My argument is with his interpretation of his claim that

a basic understanding of how language works should be part of what every educated person knows

I have a fundamental problem with the very notion of ‘educated person’ because of its pernicious political baggage. But in this post I’ve used it to accept the basic premise that if we’re going to teaching lots of stuff to children in schools, we might as well teach them the good stuff. Perhaps, not always the most immediately useful stuff but definitely the stuff that reflects the best in what we have to offer to ourselves as the humanity we want to be.

But if that is the case, then I don’t think his offer of

a modern version of the old-fashioned idea that grammar (and logic and rhetoric :-)) should be for everyone

is that sort of stuff. Let’s look at what that kind of education did for the likes of Orwell, and Stunk and White who have had the benefit of all the grammar education a school master’s cane can beat into a man and yet committed such outrageous, embarrassing and damaging transgressions against linguistic knowledge (not infrequently decried on the Language Log).

The point is that ‘grammar’ (and ‘logic’ and ‘rhetoric’) do not represent even a fraction of the principles involved in how language works. The only reason why we would privilege their teaching over the teaching of the things I propose (which cover a much larger area of how language works) is because they have been taught in the past. But why? Basing it on something as arbitrary as the hodgepodge that is the treebank terminology proposed by Mark Liberman only exposes the weakness of the argument – sure, it’s well known and universally understood by professional linguists but it hides as much about language as it reveals. And as Mark acknowledges, the aim is not to educate future linguists. There are alternatives such as Dickson’s excellent Basic Linguistic Theory that take into account much more subtly the variation across languages. But even then, we avoid all the really interesting things about language. I’m not against some very basic metalinguistic terminology to assist students in dealing with language but parsing sentences for no other reason than to do it seems pointless.

The problem with basing a curriculum on old-fashioned values is that we are catering to the nostalgia of old men (and sorry Mark, despite my profound appreciation for your work, you are an old man). (By the way, I use ‘men’ to evoke a particular image rather than to make any assertions about the gender of the person in question.) But it’s not just nostalgia. It’s also their disorientation in a changing world and discomfort with encountering people who are not like them – and, oh horror, can’t tell the passive voice from the past tense. Yes, it would be more convenient for me, if everyone I spoke to had the same appreciation for what an adverb is (particularly when I was teaching foreign languages). But is that really the best we have to offer when we want to show what should be known? How much of this is just the maintenance of the status of academics who want to see their discipline reflected in the cauldron of power and respectability that is the school curriculum? If the chemists get to waste everyone’s time with polymers, why not us with trees and sentence diagrams? In a follow up post, Dick Hudson claims that at present “we struggle to cope with the effects of [the disaster of no grammar teaching]”. But I don’t see any disaster going on at the moment. Why is teaching no grammar more disasterous than the teaching of grammar based on Latin and Greek with little connection to the nature of English? Whose rules are we after?

The curriculum is already full to bursting with too much stuff that someone threw up as a shibboleth for being educated and thus eligible for certain privileges. But perhaps our curriculum has now become the kind of stable that needs the janitorial attention of a modern Heracles.

Post script: Minimalist metalinguistic curriculum

I once analysed the Czech primary curriculum and found over 240 metalinguistic terms. I know, riddiculous. The curriculum was based on the work of eminent Czech structuralists (whose theorizing influenced much of the rest of the world). It didn’t make the Czechs any more educated, eloquent, or better at learning foreign languages – although it did make it easier for me to study linguistics. But as I said above, there is certainly some place for metalanguage in general education. Much of it comes from stylistics but when it comes to grammar, I’d stick to about 15. This is not a definitive list:

  1. Noun
  2. Verb
  3. Adjective
  4. Adverb
  5. Preposition
  6. Pronoun
  7. Prefix
  8. Suffix
  9. Clause
  10. Past form of verb
  11. Future form of verbs
  12. Present form of verbs
  13. Subject
  14. Object
  15. Passive

Languages with rich morphology might need a few others to cover things like case but overall in my career as a language educator, I’ve never felt the need for any more, and nor have I felt in the presence of uneducated people of people who couldn’t tell me what the infinitive was. In fact, I’d rather take some items away (like adverb, prefix, suffix, or clause) than add new ones.

Sentence diagramming is often proposed as a way of instilling some metalinguistic awareness. I don’t see any harm in that (and a lot of potential benefit). But only with the enormous proviso that students use it to learn the relationship between parts of their language in use and NOT as a gateway to a cancerous taxonomy pretending to the absolute existence of things that could easily be just artifacts of our metacognition.

Things are different when it comes to the linguistic education of language teachers. On the one hand, I’m all for language teachers having a comprehensive education in how language works. On the other hand, I have perpetrated a lot of such teacher training over the years and have watch others struggle with it, as well. And the effects are dispiriting. I’ve seen teachers who can diagram a sentence with the best of them and are still quite clueless when it comes to understanding how speech acts work. Very often language teachers find any such lessons painful and something to get through. This means that the key thing they remember about the subject is that linguistics is hard or boring or both.

Photo Credit: CarbonNYC via Compfight cc

Binders full of women with mighty pens: What is metonymy

Metonymy in the wild

""Things were not going well for Mitt Romney in early autumn of last year. And then he responded to a query about gender equality with this sentence:

“I had the chance to pull together a cabinet, and all the applicants seemed to be men… I went to a number of women’s groups and said, ‘Can you help us find folks?’ and they brought us whole binders full of women.” http://en.wikipedia.org/wiki/Binders_full_of_women

This became a very funny meme that stuck around for weeks. The reason for the longevity was the importance of women’s issues and the image of Romney himself. Not the phrase itself. What it showed or rather confirmed that journalists who in the same breath bemoan the quality of language education are completely ignorant about issues related to language. Saying things like:

In fairness, “binders” was most likely a slip of the tongue. http://edition.cnn.com/2012/10/17/opinion/cardona-binders-women/index.html

The answer to this is NO. This was not some ‘freudian slip of the tongue’ nor was it an inelegant phrase. It was simply a perfectly straightforward use of metonymy. Something we use and hear used probably a dozen times every day without remarking on it (or mostly so – see below).

What is metonymy

Metonymy is a figure of speech where something stands for something else because it has a connection to it. This connection can be physical, where a part of something can stand for a whole and a whole can stand for one of its parts.

  • Part for a whole: In I got myself some new wheels., ‘wheels’ stand in for ‘car’.
  • Whole for a part: In My bicycle got a puncture., ‘bicycle’ stands for a ‘tyre‘ which is a part of the it.

But the part/whole relationship does not have to be physical. Something can be a part of a process, idea, or configuration. The part/whole relationship can also be a membership or a cause and effect link. There are some subdomain instantiations where whole sets of conventional metonymies often congregate. Tools also often stand for jobs and linguistic units can stand for their uses. Materials can also be used to stand for things made from them. Some examples of these are:

  • Membership for members: “The Chess club sends best wishes.” < the ‘chess club’ stands for its members
  • Leader for lead: “The president invaded another country.” < the ‘president’ stands for the army
  • Tool  for person: “hired gun” < the tool stands for the person
  • Linguistic units for uses: “no more ifs and buts’ < if’ and ‘but’ stand for their types of questions
  • End of a process for process: “the house is progressing nicely” < the ‘house’ is the final end of a process which stands for the process as a whole.
  • Tool/position for job“chair person” < ‘chair’ stands for the role of somebody who sits on it.
  • Body part for use: “lend a hand”, the ‘hand’ stands for the part of the process where hands are used.
  • City for inhabitants: “Detroit doesn’t like this” < the city of ‘Detroit’ stands for the people and industries associated with the city.
  • Material for object made from material: “he buried 6 inches of steel in his belly” < the steel stands for a sword as in “he filled him full of lead”, lead stands for bullets.

Metonymy chaining

Metonymies often occur in chains. A famous example by Michael Reddy is

“You’ll find better ideas than that in the library.”

where ideas are expressed in words, printed on pages, bound in books, stored in libraries.

In fact the ‘binders full of women’ is an example of a metonymic chain where women stand for profiles which are written on pages contained in binders.

It has been argued that these chains illustrate the very nature of metonymic inference. (See more below in section on reasoning.) In fact, it is not unreasonable to say that most metonymy contains some level of chaining or potential chaining. Not in cases of direct parts like ‘wheels’ standing for ‘cars’ but in the less concrete types like ‘hands’ standing for help or ‘president’ for the invading army, there is some level of chaining involved.

Metonymy vs. synechdoche

Metonymy is a term which is a part of a long standing classification of rhetorical tropes. The one term from this classification that metonymy is most closely associated with is synechdoche. In fact, what used to be called synechdoche is now simply subsumed under metonymy by most people who write about it.

The distinction is:
Synechdoche describes a part standing for a whole (traditionally called pars pro toto) as in ‘The king built a cathedral.’ or the whole standing for a part (traditionally called totum pro parte) as in ‘Poland votes no’
Metonymy describes a connection based on a non-part association such as containment, cause and effect, etc. (see above for a variety of examples)

While this distinction is not very hard to determine in most cases, it is not particularly useful and most people won’t be aware of it. In fact, I was taught that synechdoche was pars pro toto and metonymy was totum pro parte and all the other uses are an extension of these types. This makes just as much as sense as any other division but doesn’t seem to be the way the ancients looked at it.

Metaphor vs. metonymy

More commonly and perhaps more usefully, metonymy is contrasted with metaphor. In fact, ‘metaphor/metonymy’ is one of the key oppositions made in studies of figurative language.

People studying these tropes in the Lakoff and Johnson tradition will say something along the lines of metonymy relies on continguity wheras metaphor relies on similarity.

So for example:

  • you‘re such a kiss ass” is a metaphor because ‘kissing ass’ signifies a certain kind of behavior, but the body part is not involved, while
  • “I got this other car on my ass” is a metonymy because ‘ass’ stands for everything that’s behind you.

Or:

  • all men are pigs” is a metaphor because we ascribe the bad qualities of pigs to men but
  • this is our pig man” is a metonymy because ‘pig’ is part of the man’s work

Some people (like George Lakoff himself) maintain that the distinction between metaphor and metonymy represent a crucial divide. Lakoff puts metonymic connections along with metaphoric ones as the key figurative structuring principles of conceptual frames (along with propositions and image schemas). But I think that there is evidence to show that they play a similar role in figurative language and language in general. For example, we could add a third sentence to our ‘ass’ opposition such as ‘she kicked his ass’ which could be either metonymic when actual kicking occured but only some involved the buttocks or metaphoric if no kicking at all took place. But even then the metaphor relies on an underlying metonymy.

When we think of metaphor as a more special instance of domain mapping (or conceptual blending, as I do on this blog), then we see that very similar connections are being made in both. Very often both metaphor and metonymy are involved in the same figurative process. There is also often a component of social convention where some types of connections are more likely to be made.

For example, in “pen is mightier than the sword” the connections of ‘pen’ to writing and ‘sword’ to war or physical enforcement is often given as an example of metonymy. But the imagery is much richer than that. In order to understand this phrase, we need to compare two scenarios (one with the effects of writing and one with the effects of fighting) which is exactly what happens in the conceptualisation taking place in metaphors and analogies. These two processes are not just part of a chain but seem to happen all at once.

Another example is ‘enquiring minds want to know’ the labeling of which was the subject of a recent debate. We know that minds often metonymically stand for thinkers as in ‘we have a lot of sharp minds in this class’. But when we hear of ‘minds’ doing something, we think of metaphor. This is not all that implausible because ‘my mind has a mind of its own’ is out there: http://youtu.be/SdUZe2BddHo. But this figure of speech obviously relies on both conceptualisations at once (at least in the way some people will construe it).

Metonymy and meronymy

One confusion, I’ve noticed is putting metonymy into opposition to meronymy. However, the term ‘meronymy has nothing to do with the universe of figurative language. It is simply a term for a name used to label a the meaning of a word in relationship to another word where one of these words denotes a whole and another its part. So ‘wheels’ are a meronym of ‘car’ and ‘bike’ but calling a nice car ‘sharp wheels’ is synechdoche, not meronymy as this post http://wuglife.tumblr.com/post/68572697017/metonymy-or-meronymy erroneously claims.

Meronyms together with hyponyms and hyperonyms are simply terms that describe semantic relationships between words. You could say that synechdoche relies on the meronymic (or holonymic) relationship between words or that it uses meronyms for reference.

It doesn’t make much difference for the overall understanding of the issues but perhaps worth clarifying.

William Croft also claims that meronymy is the only constituent relationship in his radical construction grammar (something which I have a lot of time for but not something hugely relevant to this discussion).

Metonymic imagery

Compared to metaphor, metonymy is often seen as the more pedestrian figure of speech. But as we saw in the reactions to Romney’s ‘binders of women’ that this is not necessarily the case:

he managed to conjure an image confirming every feminist’s worst fears about a Romney presidency; that he views women’s rights in the workplace as so much business admin, to be punched and filed and popped on a shelf http://www.theguardian.com/world/shortcuts/2012/oct/17/binders-full-of-women-romneys-four-words

The meme that sprang up around it consisted mostly of people illustrating this image, many of which can be found on http://bindersfullofwomen.tumblr.com (see one such image above).

This is not uncommon in the deconstructions and hypostatic debates about metonymies. ‘Pen is mightier than the sword’ is often objected to on the basis that somebody with a sword will always prevail over somebody with a pen. People will also often critique the ’cause of’ relationships, as in ‘the king did not erect this tower, all the hard-working builders did’. Another example could be all the gruesome jokes about ‘lending a hand’ or ‘asking for a hand in marriage’. I still remember a comedy routine from my youth which included the sentence, “The autopsy was successful, the doctor came over to me extending a hand…for me to take to the trash.”

But there is a big difference in how the imagery works in metonymy and metaphor. Most of the time we don’t notice it. But when we become aware of the rich evocative images that make a metaphor work, we think of the metaphor as working and those images illustrate the relationship between the two domains. But when we become aware of the images that are contained in a metonymy (as in the examples above), we are witnessing a failure of the metonymy. It stops doing its job as a trope and starts being perceived as somehow inappropriate usage. But metaphor thus revealed typically does its job even better (though not in all cases as I’ve often illustrated on this blog).

Reasoning with metonymy

Much has been written about metaphoric reasoning (sometimes in the guise of ‘analogic reasoning’) but connection is just as an important part of reasoning as similarity is.

Much of sympathetic magic requires both connection and similarity. So the ‘voodoo doll’ is shaped like a person but is connected to them by a their hair, skin, or an item belonging to them.

But reasoning by connection is all around us. For instance, in science, the relationship of containment is crucial to classification and much of logic. Also, the question of sets being part of sets which has spurred so much mathematical reasoning has both metaphoric and metonymic dimensions.

But we also reason by metonymy in daily life when we pay homage to the flag or call on the president to do something about the economy. Sometimes we understand something metonymically by compression, as if when we equate the success of a company with the success of its CEO. Sometimes we use metonymy to elaborate as when we say something like 12 hard working pistons brought the train home.

Metonymy is also involved in the process of exemplars and paragons. While the ultimate conceptualization is metaphoric, we also ask that the exemplar has some real connection. Journalists engage the process of metonymy when they pick someone to tell their story to exemplify a larger group. This person has to be both similar and connected to engage the power of the trope fully. On a more accessible level, insults and praise often have a metonymic component. When we call someone ‘an asshole’ or ‘a hero’, we often substitute a part of who they are for the whole, much to the detriment of our understanding of who they are (note that a metaphor is also involved).

Finally, many elements of representative democracy rely on metonymic reasoning. We want MPs to represent particular areas and think it is best if they originate in that area. We think because we paid taxes, the police ‘work for us’. Also, the ideology of nationalism and nation states are very much metonymic.

Warning in conclusion

I have often warned against the dangers of overdoing the associations generated by metaphors. But in many ways metonymy is potentially even more dangerous because of the magic of direct connection. It can be a very useful (and often necessary) shortcut to communication (particularly when used as compression) but just as often it can lead us down dangerous paths if we let it.

Background

This post is an elaboration and reworking of my comment on Stan Carey’s post on metonymy:  It seemed to me a surprisingly confused and unclear about what metonymy does. This could be because Stan is no linguistic lightweight so I have expected more. But it’s easy to get this wrong, and rereading my comment there, it seems, I got a bit muddled myself. And, I’m sure even my more worked out description here could be successfully picked over. Even Wikipedia, which is normally quite good in this area, is a bit confused on the matter. The different entries for synechdoche and metonymy as well as related terms seem a bit patched together and don’t provide a straightforward definition.

Ultimately, the finer details don’t matter as long as we understand the semantic field. I hope this post contributes to that understanding but I’ll welcome any comments and corrections.

Three books of the year 2013 and some books of the century 1900-2013

I have been asked (as every year) to nominate three books of the year for Lidové Noviny (a Czech paper I contribute to occasionally). This is always a tough choice for me and some years I don’t even bother responding. This is because I don’t tend to read books ‘of the moment’ and range widely in my reading across time periods. But I think I have some good ones this time. Additionally, Lidové Noviny are celebrating 120 years of being a major Czech newspaper so they also asked me for a book of the century (since 1900 till now). It makes no sense to even try to pick ‘the one’, so I picked three categories that are of interest to me (language, society and fiction) and chose three books in each.

The Colorful Library of an Interaction Designer (Juhan Sonin) / 20100423.7D.05887.P1 / SMLCreative Commons License See-ming Lee via Compfight

Three books of 2013

Thanks to the New Books Network, I tend to be more clued in on the most recent publications, so 2 of my recommendations are based on interviews heard there.

A Cultural History of the Atlantic World, 1250-1820 by John K. Thornton is without question a must read for anyone interested in, well, history. Even though he is not the first, Thornton shows most persuasively how the non-Europeans on both sides of the Atlantic (Africa and the Americas) were full-fledged political partners of the Europeans who are traditionally seen simply as conquerors with their dun powder, horses and steel weapons. Bowerman shows how these were just a small part of the mix, having almost no impact in Africa and playing a relatively small role in the Americas. In both cases, Europeans succeeded through alliances with local political elites and for centuries simply had no access to vast swathes of both continents.

Raising Germans in the Age of Empire: Youth and Colonial Culture, 1871-1914 by Jeff Bowersox. This book perhaps covers an exceedingly specific topic (compared to the vast sweep of my first pick) but it struck a chord with me. It shows the complex interplay between education, propaganda and their place in the lives of youth and adults alike.

Writing on the Wall: Social Media – the First 2,000 Years by Tom Standage. Standage’s eye opening book on the telegraph (The Victorian Internet) now has a companion dealing with social communication networks going back to the Romans. Essential corrective to all the gushing paradigm shifters. He doesn’t say there’s nothing new about the Internet but he does say that there’s nothing new abou humans. Much lighter reading but still highly recommended.

Books of the Century

This really caught my fancy. I was asked for books that affected me, but I thought more about those that had an impact going beyond the review cycle of a typical book.

Language

Course in General Linguistics by Ferdinand de Saussure published in 1916. The Course (or Le Cours) Published posthumously by Saussure’s students from lecture notes is the cornerstone of modern linguistics. I think many of the assumptions have been undermined in the past 30-40 years and we are ripe for a paradigm change. But if you talk to a modern linguist, you will still hear much of what Saussure was saying to his students in the early 1900s in Geneva. (Time to rethink the Geneva Convention in language?)

Syntactic Structures by Noam Chomsky published in 1957. Unlike The Course, which is still worth reading by anyone who wants to learn about language, Syntactic Structures is now mostly irrelevant and pretty much incomprehensible to non-experts. However, it launched the Natural Language Processing revolution and its seeds are still growing (although not in the Chomskean camp). Its impact may not survive the stochastic turn in NLP but the computational view of language is still with us for good and for ill.

Metaphors We Live By by George Lakoff and Mark Johnson published in 1980 while not completely original, kickstarted a metaphor revival of sorts. While, personally, I think Lakoff’s Women, Fire, and Dangerous Things is by far the most important book of the second half of the 20th century, Metaphors We Live By is a good start (please, read the 2003 edition and pay special attention to the Afterword).

Society

The Second Sex by Simone de Beauvoir published in 1949 marked a turning point in discourse about women. Although the individual insights had been available prior to Beauvoir’s work, her synthesis was more than just a rehashing of gender issues.

Language and Woman’s Place by Robin Tolmach Lakoff published in 1973 stands at the foundation of how we speak today about women and how we think about gender being reflected in language. I would now quible with some of the linguistics but not with the main points. Despite the progress, it can still open eyes of readers today.

The Savage Mind by Claude Levi-Strauss published in 1962 was one of the turning points in thinking about modernity, complexity and backwardness. Strauss’s quip that philosophers like Sartre were more of a subject of study for him than valuable interlocutors is still with me when I sit in on a philosophy seminar. I read this book without any preparation and it had a profound impact on me that is still with me today.

Fiction

None of the below are my personal favourites but all have had an impact on the zeit geist that transcended just the moment.

1984 by George Orwell published in 1949. Frankly I can’t stand this book. All of its insight is skin deep and its dystopian vision (while not in all aspects without merit) lacks the depth it’s often attributed. There are many sci-fi and fantasy writers who have given the issue of societal control and freedom much more subtle consideration. However, it’s certainly had a more profound impact on general discourse than possibly any piece of fiction of the 20th century.

The Joke by Milan Kundera published in 1967 is the only book by Kundera with literary merit (I otherwise find his writing quite flat). Unlike Orwell, Kundera has the capacity to show the personal and social dimensions of totalitarian states. In The Joke he shows both the randomness of dissent and the heterogeniety of totalitarian environments.

The Castle by Franz Kafka published in 1912 (or just the collected works of Kafka) have provided a metaphor for alienation for the literati of the next hundred years. I read The Castle first so it for me more than others illustrates the sense of helplessness and alienation that a human being can experience when faced with the black box of anonymous bureaucracy. Again, I rate this for impact, rather than putting it on a ‘good read’ scale.

My personal favorites would be authors rather than individual works: Kurt Vonnegut, Robertson Davies, James Clavell would make the list for me. I also read reams of genre fiction and fan fiction that can easily stand up next to any of “the greats”. I have no shame and no guilty pleasures. I’ve read most of Terry Pratchett and regularly reread childhood favorites by the likes of Arthur Ransome or Karl May. I’ve quoted from Lee Child and Tom Clancy in academic papers and I’ve published an extensive review of a Buffy the Vampire Slayer fan fiction novel.

Finally, for me, the pinnacle of recent literary achievement is Buffy the Vampire Slayer. I’ve used this as an example of how TV shows have taken over from the Novel, as the narrative format addressing weighty issues of the day, and Buffy is one of the first examples. Veronica Mars is right up there, as well, and there are countless others I’d recommend ‘reading’.

Hacking Metaphors, Frames and Other Ideas