Re: PK epistemology

[Colin Danby <danbyc@xxxxxxx>]

Dear Ed,

Thanks much for  your response.  I’m going to be long, with apologies.

I don’t think PK is inherently or innately or incurably functionalist, at
least as a tradition of social analysis.  I do think functionalist arguments
have been deployed by a number of PK writers for a variety of reasons, and
that these arguments, and the assumptions they draw in, needlessly limit
the scope of PK analysis.  Let me take a moment to pin this point down. 
Here are Dow’s five elements of PK methodology:

> 1. The economic process is seen as being ordered to a considerable
> degree because of institutions and conventional behavior, rather than market
> coordination.  However, there is always the possibility for disorder due
> to creative, profit-seeking behavior and the revision of expectations under
> conditions of uncertainty. Economic structures are not inherently stable,
> either because institutions and conventions may change, or because expectations
> may change discretely or the degree of uncertainty itself may change.
>   
> 2. There are inevitable limits to knowledge.  This means that, in general,
> knowledge of economic relations is held with uncertainty.  Uncertainty is
> the norm since social systems transform themselves unpredictably.
>   
> 3. Money is the social institution devised to address uncertainty.  It serves
> as the denominator of labor and debt contracts, and allows capitalist economies
> to function. 
>   
> 4. There is a focus on production rather than exchange as the driving force
> in the economy.  Demand and supply are understood as interdependent, and
> imperfect competition is seen as the normal structure in product and factor
> markets. 
>   
> 5. Income distribution and, more generally, the distribution of economic
> power are perceived as significant social issues for economics to address.
>   
>   
> (Dow 2001, 19)

Note how differently (3) is phrased as compared to the rest.  I read Dow’s
(3) as functionalist.  (It assumes a single need that prompts the emergence
of observable monies, it seems to suggest that only money can fill this need,
it assumes that money or monetary institutions meet this need well, and it
also biases the analysis toward a single standard variety of money or monetary
institutions.)  I would accept 1, 2, 4, and 5 but replace (3) with:

> 3'. Material processes occur in irreversible historical time,
> and we expect that, in any particular place or time, a variety of social
> institutions enable and/or constrain the time-arrangement and time-coordination
> of those material processes.  

My language of “enable or constrain” is intended to keep the notion that
institutions do important things and have important effects without assuming
that that is why they are there (or that they even do those things especially
well).  (Those who have read Stephen Jay Gould on spandrels will have a better
sense of what I mean.)  I acknowledge that my proposed substitution has various
consequences, which we could pursue.

As I read over your post, the key phrase appears to be “accounts for,” in
the sense that “accounting for” the existence of an institution appears to
drive the rest of the argument.  Perhaps it might be worth dwelling on that
before diving into more specifics.  To make a quick-and-dirty schema of five
kinds of social ontology and explanation:

> A. Pure contingency, accident, fluidity, complete freedom to
> remake the world etc. 
>   
> B. Starting with existing institutions and trying to map them, understand
> their workings, and trace their effects.  Here one might “give an account
> of” the various effects of an institution, what it does – in my preferred
> language, the specific material practices it enables or constrains and how
> it enables or constrains them.  This includes the possibility that actually-observed
> institutions may have a certain shagginess and unevenness and may at certain
> points bleed into other institutions.  (I’m thinking about the way, for example,
> Jacques Donzelot and Jacqueline Stevens have argued that observable kinship
> in specific places and times is in part a state institution for regulating
> individuals.)
>   
> C. Explaining the rise and fall of observable institutions in actual history,
> which allows for a certain amount of contingency and accident, various path-dependencies
> etc.  E.g. if I were going to try to account for the British House of Lords,
> the Napoleonic Code, or the differences between the Mexican and U.S. systems
> of higher education, I would inevitably bring in a variety of much larger
> forces and not just rely on the inherent properties of those institutions. 
> In actual history highly useful and desirable institutions may be wiped out
> by other forces; silly or ugly ones imposed.
>   
> D. Functionalism, which postulates a single originary need, a sort of Ursprung
> that in some logical and historical way generates the observed institution.
>   
> E. Arguments about the inevitability of specific institutions (essentially,
> functionalism plus a closed world).

We both find (A) and (E) unpersuasive.  I like (B) and (C), and I doubt you’re
resistant to them as stated.  (I think they're also consistent with Fred
Lee's vision.)  I think (B) and (C) could support the *full* range of PK
insights into the observable world (and more!), at the same time that we
forswear (D).   I would see (B) and (C) as “giving an account of” but would
bracket off “accounting for” an institution, which may not actually be possible. 
(At the least it seems to me to postulate some higher plane of social existence
where unmediated needs live.)

To return briefly to proposition 3 in Dow above.  If we have to speak of
observed money as fulfilling functions, and if we have to root causal explanations
back into just-so stories of money’s emergence, do we needlessly limit our
explanations?  Suppose an aspect of money in a certain situation is a spandrel,
in Gould’s sense – a presently useful thing that emerged for very different
reasons (there’s actually some of this kind of explanation in early Keynes). 
Additionally, Dow’s proposition 3 tends to put aside the possibilities that,
say, kinship or other aspects of the state or other institutions entirely
may structure material life in ways that “address uncertainty.”  (Dow’s propositions
(1) and (3) almost appear to partition institutions into those making uncertainty
and those resolving it)  We then also rule out insights that may come from
different kinds of kinship systems or different kinds of government organizations
in specific places and times.  

I see Ted has already made some points about the 1-5 sequence in your last
post, so I’ll keep my comments on it to a minimum.  I do find the argument
in your last post at least proto-functionalist, in the sense that it could
be a preamble to specifically functionalist arguments.  

> (i) I accept your (1) and (2).  The key logical turn happens
> in your (3), which counterposes openness and structure.  On the one hand
> openness, which has just been linked to uncertainty in (2), acquires a negative
> connotation (worry? doubt? excessive fluidity?).  On the other hand “structure”
> is brought in as a good thing that lets people “lead a meaningful life.” 
>   
>   
> (ii) When we observe people, they are always already in structure(s), and
> always already in “meaning” in at least the linguistic and cultural senses. 
> Being human, as opposed to being a squirrel or a mayfly, is to be living
> within meaning.  So simply to add a “need and a desire to lead a meaningful
> life” as a postulated origin or causal root for existing institutions and/or
> structures does not, it seems to me, add explanatory power.  Plus we are
> clearly phrasing “meaningful life” broadly enough to make your (3) and (4)
> unfalsifiable.  Nor does this strike me, phrased as it is, as adding to our
> ability to explain change.  Since observed people always already have meaning
> and structure, why would they opt for new systems of meaning and/or structure
> and/or scrap old ones?  
>   
> (iii) The argumentative sequence you provide appears to rule out an alternative
> tradition of structural explanation which is that some institutions may be
> means for one group of people to control and/or exploit another.  Since I’m
> arguing against “accounting for” stories I can’t argue for the merits of
> this alternative account, but I raise it as a contrast to point out the Whiggish
> implications of the argument you set out – it may bias us toward understanding
> what we see as helping people live meaningful lives.  

Best, Colin


Dow, Sheila. (2001) “Post Keynesian Methodology” pages 11-20 in Richard P.
Holt and Steven Pressman, eds., _A New Guide to Post Keynesian Economics_.
London: Routledge.