Tuesday, March 26, 2013

Currencies can be 'ultimate' or 'proximal'

In earlier postings I’ve promised some attempts at clarifying and tidying up talk about common currencies. Here is a first contribution to that project.

Generically, and of necessity vaguely because different more precise formulations appropriate to specific sciences pull in (partly) incompatible directions:

A common currency is a scalar quantity of which different options (rewards, punishments, behaviours, actions, commodities) have varying amounts. This may apply to all options, or merely to those that are in some sense “in play” for an agent at a given time.

How can currency talk be made less equivocal?

The first distinction that I need to draw is between currencies that represent some kind of order that is manifest in the choices an agent makes, and currencies that represent some aspect of the processes by which individual choices are actually produced.

Appropriating – with modest tweaks to be signaled below – some terms from a cognate area, I will call the first kind of currency ‘ultimate’ and the second ‘proximal’.

An ultimate currency is a construct in a descriptive or explanatory theory of the behavior of some agent. An ultimate currency assigns values to behaviours, or to what those behaviours achieve or (perhaps) have the function of achieving.

(We might also call this kind of currency ‘theoretical’ or ‘behavioural’ or ‘attributed’, although all of these options bring at least some unwelcome or unhelpful associations.)

Examples of ultimate currencies include various behavioural ecological versions where value is a function of fitness, and behavioural psychological versions expressed in terms of reward and reinforcement, as well as microeconomic versions that refer to utility of some kind. Considered as a set of interpretive facts, folk-psychological ascriptions of desires are also candidates for ultimate currencies.

A proximal currency, on the other hand, contributes to the process of behaviour/action selection by an agent.

(We might also call such a currency ‘causal or ‘productive’ or ‘mechanical’, although here too not all of the associations are helpful.)

Examples of proximal currencies include neuroeconomic models of the representation of subjective value in the brain.

Claims about proximal currencies usually include claims about ultimate currencies, but the converse is not generally true. The main reason that claims about proximal currencies usually include claims about ultimate currencies is that orderly choice (in some sense of ‘orderly’) is taken to follow from order in the decision making process, and so a currency in the latter should be manifest in the former. The main reason that the converse is not generally true is that at least some claims about ultimate currencies are offered in a spirit of (at least) agnosticism about causal explanation. Many micro-economists, for example, self-consciously exclude considerations of psychological implementation from their work. And some ultimate currencies (for example, among those who hold that at least some plants are capable of behavior, and that their behavior is adaptive) are described for agents that have no cognitive states at all.


  • This is a preliminary and rough statement of one important distinction. I’ll develop this distinction more fully in later postings, and also add further refinements by distinguishing sub-types of both ultimate and proximal currencies.
  • I’ll also add commented lists of examples of at least some sub-types, so that the classification takes on more substance.
  • The lists of examples (once the sub-types have been explained) might turn out to be the subjects of further evidence hooks.

Future postings:

  • Ultimate currencies: Evolutionary vs. Subjective
  • Proximal currencies: Cognitive vs. Somatic
  • List of Subjective Ultimate currencies
  • List of Cognitive Proximal currencies

Monday, March 25, 2013

Brief notices 1

I’ve decided to post periodic brief notices about things relating to the Common Currency topic, but where I’ve not yet developed a more detailed response. (So in most cases the idea is that in some future posts I’ll say something more comprehensive.)

First, here’s an interesting and recent review article (in an Open Access journal) by Peter Shizgal:

Shizgal P (2012) Scarce means with alternative uses: Robbins’ definition of economics and its extension to the behavioral and neurobiological study of animal decision making. Front. Neurosci. 6:20. [Journal link - open access] [doi: 10.3389/fnins.2012.00020]
Few have done more than Peter Shizgal to make a certain part of the empirical case that some species exhibit patterns of choice for which a central and unified value representation (that is, a Common Currency) is a good, or perhaps the only, explanation. I’ll discuss some of his empirical work at more length in the future. In this review article he makes the case for using Robbins’ definition of economics to inform the study of animal decision-making.

Second, I read an interesting op-ed piece by Scott Atran in Foreign Policy [Access may be subject to annoying pop-ups].

The main point of the article is to argue that immanent cuts to social science funding threaten national security. In the course of making this point, he glosses some of his current work relating to sacred values and human conflict. The first passage that specifically struck my eye was this one:
“Humans process these identities as moral rules, duties, and obligations that defy the utilitarian and instrumental calculations of realpolitik or the marketplace. Simply put, people defending a sacred value will not trade its incarnation (Israel's settlements, Iran's nuclear fuel rods, America's guns) for any number of iPads, or even for peace.”
Shortly after that, he makes clear that he both knows and means what he’s saying here:
“The sacred values of "devoted actors," it turns out, generate actions independent of calculated risks, costs, and consequences -- a direct contradiction of prevailing "rational actor" models of politics and economics, which focus on material interests. Devoted actors, in contrast, act because they sincerely and deeply believe "it's the right thing to do," regardless of risks or rewards.” [emphasis added.]
I’m not decided yet whether this interpretation of his evidence is the best one. But there is indeed a growing body of evidence (both from his work, and that of other social scientists). You can see some of his related papers here. All I want to note for now is that this looks like prima facie documentation of decision problems that suggest there is not a single common currency. That is, the alleged impossibility of paying some people to act contrary to their sacred values, or the ways that their actions in service of those values are insensitive to risk and reward, suggest that those values just don’t appear on the same scale as other things they will work for.

Saturday, March 23, 2013

Important Sources: Kim Sterelny

Book Cover: Thought in a Hostile World
I’m currently reading Kim Sterelny’s Thought in a Hostile World (Blackwell/Wiley 2003). Part of why I’m reading it is for the purposes of this common currency project. (The book as a whole is profoundly interesting and worth reading.)

Thought in a Hostile World is concerned with two related projects concerning the cognitive evolution of humans. One is about the ‘wiring and connection facts’ from the perspective of a theory of human agency, and evolutionary history. The other concerns the relationship between the wiring and connection facts, and a second set of ‘interpretive’ facts. These interpretive facts (about humans) include belief-desire folk psychology, but also sciences such as economics and anthropology which work with what Sterelny calls “refined versions of our folk self-conception”.

With respect to belief Sterelny argues on the one hand that many examples of successful control systems can get by with less than beliefs, and on the other than some specific types of cognitive demand would explain the evolution of belief-like states, because they require ‘de-coupled’ representations. These representations encode information that cannot be reliably extracted from single cues, that are potentially relevant to many actions, and which may be relevant at times other than when the information is perceptually salient.

After three chapters mostly focused on belief, and belief-like states, Sterelny devotes on (Chapter 5, pages 78-96) to “The Descent of Preference”. I don’t yet have a settled view about what Sterelny has to say here. But it is definitely relevant and important for anyone thinking common currencies for action selection:
  • Although making no direct reference to common currencies, and surprisingly little to the economic aspects of choice, Sterelny is concerned with the natural history of preference, and with the question how an agent with a preference structure of some kind (he does occasionally use the expression ‘utility function’) might behave differently from one without.
  • He argues that preferences are like belief-like states (‘decoupled representations’) in being found only in some animals, and in response to specific cognitive challenges that cannot be solved by simpler means. Sterelny is (it seems) agnostic about whether rats exhibit the required sophistication, but thinks that primates definitely do.
  • So, depending on what exactly Sterelny means by preferences, his view is some kind of partial eliminativism about preferences (some animals may seem to have them, but if you use the right criteria you’ll see that they don’t need them). It would seem to be a consequence of this position that common (cognitive) currencies are relatively rare and unusual.

Part of the reason I don’t yet have a settled view about Sterelny’s chapter on “The Descent of Preference” is that I’m currently downright puzzled by some of the decisions he seems to have taken in the chapter. Among them:

  • The companion to belief in intentional psychology is desire. In previous chapters he’s been careful to explain how belief-like states and decoupled representations are pretty close to beliefs, without (very reasonably) calling for them to be identified. Yet in this chapter he mostly talks about preference as opposed to desire. It’s consequently less clear what explanatory target he’s aiming at.
  • This is made worse by some of the things he says about preferences. So at one point he describes the task of the chapter as giving an account of “motivation based on representations of the external world.” Beliefs are paradigmatically representational states. And desires are supposed to interact with beliefs. But why make desire itself (or preference) primarily representational? The main theories of desire on the market are mostly non-representational. They have it that desire is a disposition to action, or has something to do with pleasure, attention, learning. (There are fairly representational theories of desire too, but they’re hardly the most congenial to Sterelny’s natural history of cognition.)
  • Sterelny organizes some of his discussion by reference to what calls ‘the forager’s dilemma’ (which he does not clearly define in the chapter). But he says surprisingly little about what I above called the ‘economic aspects of choice’. I have in mind here the problem Robbins (1935) formulated as that of allocating “scarce means which have alternative uses”. Any agent with a relatively rich behavioral repertoire faces a trade-off problem in allocating resources between available actions. Many actions are mutually exclusive, either directly or in drawing from a common energy budget. The costs of far-from optimal allocation might be very high. I haven’t yet figured out why Sterelny says relatively little about this kind of issue.

Still, it’s a fine book and well worth reading. I’ll develop some more systematic reflections on it in due course, and post them here where they can be safe from prying eyes. (I'll also comment on some of the earlier papers in which Sterelny developed some of the ideas that appear in Thought in a Hostile World.)


Robbins, L. (1935) An Essay on the Nature and Significance of Economic Science. London: MacMillan and Co.

Sterelny, K. (2003) Thought in a Hostile World. Oxford: Blackwell.

Tuesday, March 19, 2013

My own efforts

This page is an overview of the development of my own attempts to contribute something to the common currency topic. It will be periodically updated as progress actually gets made. Where appropriate I'll add links to preprint or published versions of any papers. If you would like to receive a copy of a paper when it's ready to be circulated, email me or announce the fact in the comment thread below.

Talks and Conference presentations

"The Natural History of Desire"

This is my presentation for the 2015 conference of the Philosophical Society of Southern Africa (PSSA). I did a rough pilot version at a smaller conference in South Africa in September 2013, but this is rather more worked out. I also presented an abbreviated version at the European Society for Philosophy and Psychology in Estonia in 2015. For now, the link above takes you to the abstract - a working text of the paper will follow.

"Philosophers should be interested in 'common currency' claims in the cognitive and behavioural sciences."

This paper (presented first in January 2014) is a new attempt to introduce the common currency topic, and argue that it is warrants philosophical attention. This came into being because the introductions to the talks I was giving kept getting longer, and eventually I came to think that they were close to deserving to stand alone. We will see. Here is a working conference-paper length (about 6000 word) draft of that paper.

"Need there be a common currency for decision making?"

This was a working talk, that I gave a few times, including at the Spring Philosophy Colloquium, Grahamstown (September 2008), at the 35th annual conference of the Philosophical Society of Southern Africa, Hogsback (January 2009), and at a one day philosophy of science colloquium in Johannesburg (January 2010). Let me be frank - the talk wasn't very good, partly because I mistakenly thought I could cover a lot more ground than I actually could in a single session. There's a published version, which also isn't very good.

"Intragenomic conflict and common currencies."

This talk was presented at the annual conference of the Philosophical Society of Southern Africa, Salt Rock KZN (January 2013), and at a few departmental colloquia. The written version (as opposed to a set of slides and notes) is coming along nicely, and I'll post a draft version of it here as soon as it's fit for public consumption. The talk focuses on some arguments made against the common currency thesis by the biologist David Haig. (Some of the papers in question are available on his website here.)
The abstract of the paper I'm writing, with the term 'common currency' dropped from the title, is here: Intragenomic conflict and Intrapersonal conflict.

"Subsumption architectures and common currencies."

I've only presented this talk in departmental occasional seminars so far, but will take it to conferences in due course. The paper attempts to consider how defenders of a common currency might respond to the anti-representational arguments (in artificial intelligence) associated with various figures, and exemplified by the early 1990s work of Rodney Brooks. Brooks isn't directly concerned to say things about utility or preference - most of his anti-representationalist fire is aimed at world-modelling. But if there's no central system, then there's no place for a common currency to be represented either.
Update: As of late 2013 this paper is going forward as a collaboration with Blaize Kaye, who is planning PhD level research on the common currency topic.

"What does neuroeconomics reduce, and to what?"

I've presented this at a few departmental occasional seminars, and at at a one day Philosophy of Science Colloquium, Salt Rock KZN (January 2013). The talk is mostly about problems with Paul Glimcher's discussion of reductionism in his generally very fine book Foundations of Neuroeconomic Analysis (Oxford, 2011).

"Inner Conflict"

This was a talk I gave a TEDxUmhlanga in September 2013. It is broadly relevant to the concerns of this blog, and draws on ideas articulated in detail in the work of George Ainslie

Working papers

If the paper title in this list is a hyperlink, it will direct you to a page dedicated to that working paper on this blog.

Published papers and commentaries

The first paper to come out of this project isn't particularly good. (It's in a conference proceedings volume, and I wouldn't encourage anyone to read it.) Others can be found here:

Three commentaries that I've written (one with a co-author) are indirectly relevant to the currency project:

Version history:

First posted March 19, 2013.
Updated May 6, 2013 (added reference to 'cui bono?'). Updated May 16, 2013 (added link to 'On the motivational commensurability of pleasure and pain'). Updated January 11, 2014 (added new entry, and links to 'Philosophers should be interested...'). Updated July 11, 2014 (updated links). Updated January 9, 2015 (added new entry).

Monday, March 18, 2013

Neural processing of rewards in different modalities takes place in the same way (in humans) – An Evidence hook

See ‘notes’ below for what is an evidence hook?

Chib, V.S., Rangel, A., Shimojo, S. and O’Doherty, J.P. (2009) Evidence for a Common Representation of Decision Values for Dissimilar Goods in Human Ventromedial Prefrontal Cortex. The Journal of Neuroscience. Vol 29(39) pp.12315–12320. [Publisher link - may be behind paywall] [DOI: 10.1523/ JNEUROSCI.2575-09.2009]
Abstract: To make economic choices between goods, the brain needs to compute representations of their values. A great deal of research has been performed to determine the neural correlates of value representations in the human brain. However, it is still unknown whether there exists a region of the brain that commonly encodes decision values for different types of goods, or if, in contrast, the values of different types of goods are represented in distinct brain regions. We addressed this question by scanning subjects with functional magnetic resonance imaging while they made real purchasing decisions among different categories of goods (food, non-food consumables, and monetary gambles). We found activity in a key brain region previously implicated in encoding goal-values: the ventromedial prefrontal cortex (vmPFC) was correlated with the subjects’ value for each category of good. Moreover, we found a single area in vmPFC to be correlated with the subjects’ valuations for all categories of goods. Our results provide evidence that the brain encodes a “common currency” that allows for a shared valuation for different categories of goods.
Comments: Common neural activity for different categories of rewards (food, non-food consumables, monetary gambles). 

McNamee, D. Rangel, A., and O’Doherty, J.P. (2013) Category-dependent and category-independent goal-value codes in human ventromedial prefrontal cortex. Nature Neuroscience, advance on line publication. [Readcube version - I'm not sure whether there's a risk of a paywall[doi:10.1038/nn.3337]
Abstract: To choose between manifestly distinct options, it is suggested that the brain assigns values to goals using a common currency. Although previous studies have reported activity in ventromedial prefrontal cortex (vmPFC) correlating with the value of different goal stimuli, it remains unclear whether such goal-value representations are independent of the associated stimulus categorization, as required by a common currency. Using multivoxel pattern analyses on functional magnetic resonance imaging (fMRI) data, we found a region of medial prefrontal cortex to contain a distributed goal-value code that is independent of stimulus category. More ventrally in the vmPFC, we found spatially distinct areas of the medial orbitofrontal cortex to contain unique category-dependent distributed value codes for food and consumer items. These results implicate the medial prefrontal cortex in the implementation of a common currency and suggest a ventral versus dorsal topographical organization of value signals in the vmPFC.

Notes: This paper seems to raise the evidential bar appreciably over previous work, including the Chib et al (2009) paper listed above. I’d be rather surprised if this paper did not turn out to represent the ‘state of the art’ on this issue for the first half of 2013. This paper defends a far more narrow and specific claim than many others that will appear on this list (or ‘hook’), and I’ll discuss it in more detail later, as well as include it in the hooks for narrower claims about neural valuation in humans.


Entries on this page will be to papers specifically claiming to show that the neural processing of rewards in different modalities (in humans) either does or does not take place 'in the same way'. This formulation is deliberately vague so that I can cast the net wide. As I go along I'll identify more restricted theses, and document them separately.

This is the first of a collection of 'evidence hooks'. Together the hooks will form an evidence rack. (What I’ll do here is inspired by a suggestion of Peter Suber’s – see this posting on Common Currencies, and this document by Peter Suber.)

I’m going to figure out how to work with the evidence hook/evidence rack idea as I go along, and it may take a while before I’ve settled on stable procedures.

Version history

First version posted - 2013/03/18.
Minor update - 2013/03/22.