Monochrome swans

I have a request[^1] for help from scientifically literate readers. A lot of my research work is focused on the problem of unforeseen contingencies, popularly, if ethnocentrically, described as “black swans”. In particular, I’m interested in the question of how you can prepare for such contingencies given that, by definition, you can’t foresee exactly what they will be. One example, with which I’m very pleased, is that of the precautionary principle. It seems reasonable to say that we can distinguish well-understood choices involving hazards from those that are poorly understood, and avoid the latter, precisely because the loss from hazard cannot be bounded in advance.

Anyway, I was thinking about this in relation to the actual case of black swans (or, from my own perspective, white swans). The question is: what principles would help you to avoid making, and acting on, the assumption “all swans are white (or, in my own case, black)”. It seems to me that the crucial fact here is that the shift from black to white, or vice versa, is, in evolutionary terms, a small one. So, if you used something like cladistics, you would avoid choosing feather color as a defining feature of swans, and birds in general. As I understand it, a phylogenetic approach starts with features that are very strongly conserved (body plans) and proceeds from there. But, rather than assume that my own understanding is correct, it seemed simpler to ask.

[^1]: There’s a blog-specific word for this, but I refuse to use it

33 thoughts on “Monochrome swans

  1. Classical logic has it that there are 2 ways of assigning probability. Empirical (from observation or experience) and a prior i (from rules and deduction). They are not totally separable from each other – but an empirical method would not predict a black swan whereas a prior i could.

    The peer review system may not be good at finding black swans as it usually ensures research unfolds methodically and cautiously in predictable directions .The big game changing revolutions often come as black swans. So many of the big scientific advances came from accidents or mistakes -even from dreams. As for black swan useful principals – I’d say ‘avoid hubris’ ,embrace doubt .

  2. For example, first work on preventing catastrophic climate change, and use the research from that to address the more general problem of getting shaved monkeys to worry about threats they can’t see that will kill them in the impossibly far distant future (viz, later than next fiscal quarter).

    I was watching a video of a talk by David Friedman the other day. He was talking about global warming and the associated temperature risk. He talked about the range of temperature increases predicted by the IPCC and then pointed out that there was a small risk that things could be much worse than predicted. He referred to this as a low probability, high cost scenario. However he then pointed out that the opposite scenario also exists. There is a low probability that without the extra CO2 we would be entering a new glacial period and that global warming is slowing or even stopping that from occurring. Under this second low probability scenario policies that stop or slow global warming are very high cost.

    Whilst you may choose to reject any notion of Swan symmetry in the particular case of global warming it would seem likely that there is in general some difficulty in knowing the symmetry of Swan occurrence in any given problem space. So in many circumstances Swans can turn up that lead you to regret not having done more of X, just as much as Swans can turn up that lead you to regret not having done less of X.

  3. Dear Prof Quiggin,

    I can probably provide some pointers to some interesting references. Apologies, if I only state the obvious…

    There are some interesting comments on Taleb’s books by Christian Robert, Larry Wasserman, Andrew Gelman, Denis Lindley and David Aldous if you want to search for them. Obviously Taleb’s style annoys lots of people. I like Aldous’ commentary the best.

    In part Taleb’s argument is anti-formalist, and anti-modeling, that is, he argues there is not much you can do and it is a bit defeatist. Although he argues that fat tail distributions help with “grey swans”.

    The most comprehensive account of decision making under uncertainty is usually considered to be due to Bruno de Finetti. Especially Prévision: ses lois logiques, ses sources subjectives, and Theory of probability. There is an excellent and very underrated book by Frank Lad, the book by Peter Walley might also be relevant.

    The most important aspects of the theory are the de Finetti representation theorem, and the fundamental theorem of prevision. The de Finetti representation provides bridges between classical and Bayesian statistics. The fundamental theorem of prevision (FTP) is less widely known, importantly it allows uncertainties to be specified imprecisely.. Bayesian statistics is currently being driven forward by computational methods that perform conditioning (or more accurately marginalization) usually using MCMC. The imprecise inference based upon the FTP is not being pursued as much because the action is elsewhere at the moment, but an imprecise probabilistic specification seems relevant to this type of ‘black swan’ problem.

    The theory importantly separates, unknown future outcomes, utility and decisions. I interpret the black swan idea as you being unable to asses the probabilities for (some) unknown future outcomes i.e. you place wide intervals on these probabilities. This in turn makes it hard to determine the expected utility of some or possibly all decisions. I expect you can motivate the precautionary principle from this sort of perspective. I would quibble a little with how you introduce the precautionary principle, in that it is the expected loss of a decision that you cannot bound – usually you have a pretty good idea of the loss of a particular decision for a particular known future outcome.

  4. «unforeseen contingencies, popularly, if ethnocentrically, described as “black swans”»

    If I understand Nassim Taleb correctly that’s not how he characterizes “black swan” events, and is it indeed very far from his conception of “black swans”.

    «In particular, I’m interested in the question of how you can prepare for such contingencies»

    I dearly hope that our poster has read the somewhat fuzzily written “Antifragile” book by Nassim Taleb. That his is reply both to “black swans” and what you are asking, except that is now ambiguous:

    «such contingencies given that, by definition, you can’t foresee exactly what they will be.»»

    OOPS, major bait-and-switch here: you have talked of “black swan”, “unforeseen contingencies” and “can’t foresee”. As to the latter two it seems to me a confusion between “unforeseen” and “unforeseeable”, where the difference between the two is pretty gigantic.

    «what principles would help you to avoid making, and acting on, the assumption “all swans are white (or, in my own case, black)”.»

    That’s a different topic again. That’s the ancient problem of extrapolation, which is based on the alternative between making sharp predictions and accepting that they may be wrong, or making vague predictions that are never wrong.

    «One example, with which I’m very pleased, is that of the precautionary principle. It seems reasonable to say that we can distinguish well-understood choices involving hazards from those that are poorly understood, and avoid the latter, precisely because the loss from hazard cannot be bounded in advance.»

    “Antifragile” by Nassim Taleb has extensive discussions of this and much more.

  5. @TerjeP

    We have a very strong idea that human activity has stopped an ice age – paleo analogues suggest that we should have been gradually moving back to glacial conditions since the holocene climatic optimum 6000 years ago.. but the small additions of CO2 from deforestation and agriculture stopped it.

    The key is in the timescales – natural ice age onset takes thousands of years, whereas the timescale for AGW is decades to centuries.

  6. @Gaius X
    Yes. JQ’s question assumes there is a connection between the problems of taxonomy of living creatures, and those of labelling events in human affairs.

    In biology, you have two problems. The one normally encountered – and the only one encountered by non-specialists – is identifying the species of an individual. Species undoubtedly exist in nature, even if there are hard cases. Jared Diamond, a card-carrying ornithologist, found that New Guinean tribesmen had the same list of bird species that he did. Only their higher classifications were different. Identifying this creature as an x, you have a lot of characters to play with. The difficulty only arises if there are very similar species. In some cases, a five-year-old can’t go wrong: giraffes, polar bears, toucans, condors. In others, you need a microscope: ants, beetles, small brown birds. But there isn’t an epistemological problem; you look it up.

    The second problem, for experts only, is fitting the species into a higher taxonomic tree. Before genetics, this was a somewhat arbitrary procedure. The higher orders were human creations, and taxonomy was not right or wrong. Cladistics tries, as I understand it,to correct taxonomy to fit a true evolutionary descent. There are I suppose difficulties for expert judgement, when different genes seem to tell different stories, or stories that don’t fit the phenotypical differences. But again, where is the epistemological problem?

    I’m guessing, but I feel that John’s economic black swan problems are closer to species identification than cladistics. That’s simple enough. Take as many characteristics as you can. Then look for a unique characteristic or two, like the toucan’s bill. (These may be environmental. A very large all-white animal in the Arctic is a polar bear). Alternatively (for small brown birds, etc), assemble a very short list of species that fit obvious characteristics, look up the differentiating characteristic, and check that.

  7. Monotremes were a surprise – egg laying, milk producing ie 2 things not previously, ethnocentrically, known to exist in the same animal and where animal groups that had one of the features had many phenotypic differences from the other. So using a larger genetic difference would not preclude black swans. Still the principle seems sound.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s