Friday, December 14, 2007

Linguists vs. Economists

CAVEAT: I have knitted this post together from random scraps of thoughts I’ve been collecting for the better part of two months. I make no apologies for its speculative nature and erratic structure. Here we go…

A rather rambling post on observing versus prescribing.

My Basic Question: What’s the cognitive difference (if any) between linguistic decisions and economic decisions?

My Other Question
: What’s the difference between the way linguists study data and the way economists study data?

As I’ve said on this blog before, over the last few years I’ve become increasingly interested in economics, particularly macro-economics. I follow several econ blogs like Greg Mankiw, Brad DeLong and Freakonomics (have not yet read the actual Freakonomics book, though)and have read a variety of popular econ books by Jeffrey Sachs and Joseph Stiglitz. For what it’s worth (not much) Mankiw can count me as a member of the Pigou Club.

But the more I read, the more uncomfortable with basic economics analysis I am becoming, and I think I’ve finally hit on why. It’s because, as a linguist, I’ve been trained in descriptive analysis, but it’s my impression that economists are in the business of prescriptions. Specifically, they prescribe optimal decision making. All economists I’ve encountered (liberal, conservative, libertarian) assume a model of human decision making where some decisions are good and others are bad (presumably based on the utility of the outcomes of those decisions). If I’ve gotten this generalization wrong, please let me know. Though I’ve never studied it formally, this seems to be the basic point of game theory.

I think it’s fair to say that few linguists think of linguistic decisions as either good or bad. Linguists avoid labeling decisions as “bad” because that would be a fundamentally prescriptivist approach. Linguists simply record decisions as facts (e.g., Americans say “ain’t” in context Y but not in context Z) then try to model the decision making as it is.

Part of the difference is probably due to the fact that economists work under the assumption that economic decisions are conscious and rational while linguists work under the assumption that linguistic decisions are automatic or unconscious; it’s not as though we think people are weighing the option between ain’t and am not then making a rational choice regarding which to choose to articulate. It is the natural human language system that makes the decision, and it’s our job as linguists to figure out what factors influence that decision. Weighing the option between a 25 cent/lb orange from California vs. a 35 cent /lb orange from Florida seems like a very different kind of decision.

This descriptive/prescriptive distinction is fundamental to studying linguistics, and it’s also the one point above all others that causes the most friction between linguists and non-linguists when discussing language issues. Popular perception holds that linguists want to change the way people talk. But this is pure fantasy. Maybe high school English teacher do, but linguists don’t. Linguists are first and foremost observers.

Can this also be said of economists? Are they trained to avoid prescriptions? My impression is, no. Quite the opposite. Economists are trained to provide prescriptions. Economists seem to be primarily concerned with policy. Policy analysis is about prescribing decisions.

For a recent example of this, see Mankiw’s post Must or Should. It seems to be the case that what Mankiw calls positive vs. normative statements, is basically the same thing as what linguist call descriptive vs. prescriptive statement.

This all started with Mankiw referencing some neuroeconomists support for the claim that the comparative size of a reward affects people’s happiness (not just the absolute size).

According to Mankiw, he and Brad DeLong have different views about people’s basic motivations (a descriptive fact). DeLong made this descriptive comment September 3, 2006:

My point was that the rich are spiteful--that they enjoy the envy of the poor.

Here’s Mankiw’s summary from Sept 3, 2006:

Brad seems to see the rich as especially mean and spiteful. I see them as some combination of more talented, hard-working, and lucky than average but otherwise like everyone else. (Or maybe Brad views everyone as mean and spiteful and the rich as having more opportunities to exercise these vile attributes.) I wonder if our varying perspectives on human nature can partly explain our different positions on public policy.

It seems to me the basic problem facing both Mankiw and DeLong (and any policy minded economist) is the challenge of teasing their respective descriptions of how people really do make decisions apart from their respective prescriptions for how governments should affect people’s decision making. Economics is fundamentally about human decision making, making it, in essence, a branch of cognitive science.

I recognize my own need to tread carefully on the ground of economics since I have no training in the area. I will gladly stipulate that economists like Mankiw and DeLong are very smart, well trained scholars who understand the complexities and nuances of economic decision making far better than the vast majority of people on this Earth (certainly far better than little ol’ me). I fear I might be reducing their analyses down to an absurd straw man out of my own ignorance of their actual, detailed work which may be far more savvy about descriptions of decision making than I am currently giving them credit for.

Nonetheless, in the spirit of off-the-top-of-your-head blogging, I wonder if they are trained in the descriptive/prescriptive distinction in the way the linguists are?

Linguists are explicitly trained to describe language as it really is (if North American English speakers say "ain't" and end their sentences with prepositions, then so be it, we study that). In fact, most linguists are trained to despise prescriptivism as inherently unnatural and counterproductive. Language is what it is.

But when I read the popular work of economists (Mankiw's blog, Sach's books, Stgilitz's book, etc) I often wonder if they are keeping the distinction straight. It seems like they confuse their prescriptions for the way people ought to make decisions with the facts of how people really make decisions (and that was the point of the neuro-economics study). Any policy suggestions linguists may recommend (e.g., anti-English only) are based on our understanding of the facts about how people really use language, not on opinion about how we feel people should use language. In other words, it seems like economic policy suggestions from Mankiw et al. are based on their (educated, intelligent) opinion about how people ought to make decisions, with the, at least implicit, hope of changing the decision making system of the average person.

I doubt any serious linguist would make policy suggestions in the hopes of changing the way people naturally make linguistic decisions.

Do economists try to sculpt economic policy suggestions to try to change the way people naturally make decisions? My hunch is, yes. And I fear this is a fool’s errand.

I have a close friend who did a B.A. in economics at Berkeley in the early nineties and graduated 3rd in his class. He went on to UCLA law school, and finally got a Harvard MBA. The guy knows his stuff. I emailed him the following questions about his economics education regarding this distinction: Were you explicitly trained to tease these two things apart? Do you recall this distinction being a principle of economic analysis? Do you think economists try to sculpt economic policy suggestions to try to change the way people naturally make decisions?

His answer:

The answer is that I don't remember much emphasis on doing things one way or the other. We clearly did want to create theories that would help people make better choices, but there were also theories about why people made bad choices where there were market break downs (generally problems with information or instances where group benefits conflict with individual benefits). Clearly though I don't remember us ever being told "do not try to change people only observe." Changing for the better would have been seen as good I think (emphasis added)

Interesting. We seem to have hit on the crux of the issue.


Erin said...

I've always been of the impression that economists do try to describe rather than prescribe, though I will agree that the bad/good terminology does seem to be more freely tossed about. Perhaps it looks different to me because the econ blogs I read are Marginal Revolution and Warren Meyer?

Chris said...

Erin, thanks for the comment. I wonder if it's a rhetorical pattern that blogs and popular econ books tend to be more prescriptive (if they are at all)? Perhaps authors see that as the point of writing a non-academic book in the first place. It's a chance to brush off the boring old descriptions and get their hands dirty with making policy suggestions. just a thought.

I'll keep up with the blogs you mentioned and see if I change my perception (I liked Cowen's post about the time it takes to get a PhD; I'm on the wrong end of that scale, hehe).

SpaceWagner said...

Hello Chris,
A friend of mine showed me your post and I was really interested on it as far as I'm concerned as an economist!
I really appreciate your analysis but I have a point here, which is the massive difference between mainstream (or orthodox) economics and alternative (heterodox) economics. You've been quoting and talking mostly about these orthodox economists, whose models are based on rational decision making, which is sth I find quite ilogical, since most of the decisions we make have a lot of influences of variables beyond the comprehension of a simple rational model. It sounds like the human being is a "lightning calculator of pleasures and pains" in these models, as would say the sociologist Thorstein Veblen.
As an heterodox economist, I would say that Economics (or I should say Political Economy) has much to improve with the contributions of other sciences to really understand the process of decision making: psychology, sociology, history, and why not, linguistics?
Thanks for your contribution! I'll probably share that into the "economics circle" lol

Chris said...

Thanks Space. I don't know much about heterodox economics. Would you consider behavioral economics as part of that? I am impressed with that growing field. Especially as a cognitive linguist, it seems to ask the right questions and create the right kinds of experiments.

Scott Sykes said...

My friend linked this to me and since I was already on his facebook page I figured I might as well put the same comment here.

"All economists I’ve encountered (liberal, conservative, libertarian) assume a model of human decision making where some decisions are good and others are bad (presumably based on the utility of the outcomes of those decisions)."
Indeed, economics does NOT stop at description. If economics stopped solely at description of what choices people do make it would be rather use-less from a macro-economic policy perspective. You must first understand how someone thinks but them from there decide what "inefficiencies" occur and how to fix them. We'll take game theory since he mentioned it and I'm in the intro class right now. In a prisoner's dilemma the equilibrium choice is for all players to defect or shirk their duty to do something (like not pollute, clean common areas, etc). This is an extremely common problem that manifests all over in our every day lives. The reason everyone shirks is that form an individual perspective, having no control over the other person, it is always the best choice. But, this is not Pareto efficient, that is all players and society as a whole can be made better off if the equilibrium is moved. This is where economists stop describing how and why people make decisions to what they should do and how to make them decide to do it of their own free will (of sorts anyway).

"Economics is fundamentally about human decision making, making it, in essence, a branch of cognitive science."
This I whole-heartedly agree with, however, as mentioned above, it can be very different from other pure sciences because it is not interested solely in "what is" but also "what could be."

"I often wonder if they are keeping the distinction straight. It seems like they confuse their prescriptions for the way people ought to make decisions with the facts of how people really make decisions (and that was the point of the neuro-economics study)."
It can sometimes be hard to entirely remove our bias on how we think the world should work, and I think this is where he really has a problem with economics. All economists have to start by describing the situation. They then move to a prescribed solution which makes society better off, given a number of assumptions. Assumptions are always the key and the crux of economics. For example, benevolent-omniscient-social planner communism has the exact same outcome as capitalism with perfect markets and information. Which then should your policies aim towards? This is where each individuals assumptions about the world and what is fair come into play and it is probably impossible to prove who is more correct.

"Do economists try to sculpt economic policy suggestions to try to change the way people naturally make decisions? My hunch is, yes. And I fear this is a fool’s errand."
I think this again lies in assumptions. There are usually inefficiencies which arise because markets are not perfect, information is asymmetric or non-existent, and we have yet to find a benevolent social planner. The goal is not to change how people decide, but to change the situation within which those choices are made, thereby changing the choice itself.

Chris said...


Thanks for the thoughtful comments. Your point that economists do not stop at descriptions is well taken, and one I hadn't fully considered. In many ways, economists get a bad reputation from other academics only because they (economists) are engaging in real world problems at a much more impactful point: policy. I don't think there are another group of academics who have a more direct say in public policy than economists. We linguists can deride all we want, but what would the world look like if the US government used Chomskyean linguistics as a basis for real-world, in-the-field language policy? I shudder at the thought.

So, you have given me pause to be more open minded. Thanks.

Scott Sykes said...

Glad I was able to paint economics in a slightly better light for you. It is definitely a joy of mine.

My linguist friend and I were talking today and it seems there are some similarities in our fields. Namely, neither works well as a pure science but it would be unfair to say the don't have quite a bit of science in them. Just a tad of philosophy, culture, and unquantifiable beliefs in both.

A linguist asks some questions about word vectors

I have at best a passing familiarity with word vectors, strictly from a 30,000 foot view. I've never directly used them outside a handfu...