Barkley,

Many thanks for reminding me of Chipman's great article and the ones  
that followed it. Chipman established without a doubt that Mill played  
a starring role in trade theory, presciently employing a startlingly  
advanced model for the time to show how the equilibrium terms of trade  
are determined. But, as you note, Mill employed no geometrical  
diagrams in his analysis, certainly not ones similar to the  
Marshallian Cross. To me this implies that he can't be considered an  
intellectual forefather of Marshall's microeconomic supply-and-demand  
graphs. I may be wrong, however.

On Feb 28, 2014, at 2:00 PM, Rosser, John Barkley - rosserjb wrote:

> Regarding Mill, while I agree that he did not draw any diagrams of S  
> and D, John SS. Chipman claimed in 1965 in "A Survey of the Theory  
> of International Trade: Part 1, The Classical Theory," Econometrica,  
> 1965, 33, 477-516, that in the 1852 3rd edition of his Principles of  
> Political Economy, Mill solved a special case of nonlinear  
> programming to verbally prove without diagrams the existence of an  
> international trade equilibrium for two countries, two goods, and a  
> special case utility function, U = X1.X2.  Later papers by James  
> Melvin in 1969 in the Southern Economic Journal and Martin Harwitz  
> in 1972 in the Journal of International Economics extended the  
> discussion, arguing that Mill's solution was more general, and  
> Chipman himself joined in considering the whole issue following  
> their discussions in 1979 in HOPE, 11, 477-500, "Mill's  
> 'Superstructure:' How well does it stand up?"
>
> ________________________________________
> From: Societies for the History of Economics [[log in to unmask]] on  
> behalf of Thomas Humphrey [[log in to unmask]]
> Sent: Thursday, February 27, 2014 11:10 PM
> To: [log in to unmask]
> Subject: Re: [SHOE] Wicksell and Stigler's Law of Eponymy
>
> Before I respond to James' latest comment, please let me reply to an
> earlier one that I failed to address.
>
> That earlier comment referred to J. S. Mill as a likely influence on,
> and even an intellectual grandfather/inventor of, Marshall's supply-
> and-demand-curve cross diagram. Contrary to James, I doubt Mill's
> relevance in this connection. For Mill never drew the Marshallian
> cross diagram, although he did discuss (in words) the concept of
> supply and demand functions. He can hardly be said to be a precursor
> -- or grandfather/inventor -- of a diagram he never drew. To jump from
> a purely verbal treatment of supply and demand to a geometrical,
> diagrammatic demonstration takes a huge leap of imagination. Mill did
> not take that leap. Marshall did, as did his predecessors Cournot,
> Rau, Mangoldt, and Jenkin. Furthermore, Mill's purely verbal
> discussion pertains not to the determination of the price of an
> individual good, but rather to the determination of the international
> terms of trade between a country's exports and its imports. In
> describing (solely in words) the determination of the relative price
> of imports in terms of exports sacrificed, Mill seems if anything to
> be more a precursor of Marshallian reciprocal demand or offer curve
> analysis than of the Marshallian S&D cross. But again, Mill drew no
> offer curve diagrams. Indeed, he drew no diagrams of any sort in his
> writings on economics. So I think Mill fails to qualify as an inventor
> of S&D geometry, or of any other geometry, for that matter.
>
> Now to James's latest comment. I'm willing to grant that IF (a big if)
> Keynes somehow managed to convince the economics profession to
> attribute the cumulative process analysis to Wicksell, Keynes did so
> inadvertently and unintentionally, not deliberately. I'm just not
> willing to believe that Keynes possessed that much power, namely the
> power to brainwash the entire economics tribe for more than 75 years..
> In short, when economists attach Wicksell's name to the cumulative
> process model, it must be because they think his name fits, not
> because Keynes told them to do so or because they were ignorant of
> earlier formulations. Wicksell's name fits because, in some key
> respects, his version is superior to, or more arresting than, earlier
> ones. His version is perceived as definitive. Perhaps Wicksell's pride
> of place stems from nothing more than his having published an entire
> book and a major article (in English, no less) on the cumulative
> process model,  whereas his predecessors published at best a few
> sentences or paragraphs on the subject. If so, then Wicksell's
> superiority lies in the completeness of his exposition. But no matter
> what the perceived strengths and advantages of his version, those are
> the strengths that must have led economists to name the model for him.
> Keynes had little to do with it.
>
> As for the claim that economists were, and are, ignorant of Wicksell's
> classical predecessors, I can't buy it. Economists understand that all
> thinkers stand on the shoulders of giants whose earlier models are
> prototypes of their own. Therefore economists must have suspected or
> known that Wicksell's cumulative process had predecessors, too. In
> this connection, I doubt that Milton Friedman was ignorant of
> Wicksell's classical forebears. Friedman was a student and later a
> colleague of Jacob Viner, Lloyd Mints, and Frank Knight, three
> scholars whose knowledge of classical economics was unmatched. It
> defies credulity that they didn't inform Friedman (and that he didn't
> discover it himself from his own reading) about classical formulations
> of the cumulative process. One surmises from these considerations that
> Friedman dubbed the analysis Wicksellian not out of ignorance of
> classical versions, but because in his estimation Wicksell's
> formulation was superior to the classical ones.