John Dewey, Embodied Cognition and Economic Theory:

Towards a Theory of Entrepreneurship


                                                                   Elias L. Khalil*


Mainstream and Marxian economic theorists have generally failed to explain entrepreneurship, and for an obvious reason.  In order to apply the calculus of optimization, neoclassical theorists have to treat the set of resources as an object that exists "in-itself," i.e., independent of the acting subject.  On the other hand, Marxian theorists, in advocating the labor theory of value, have to treat labor as abstract, i.e., can easily be measured and compared.  John Dewey, throughout his writings stretching over half a century, has undermined the objectivist view in a way which can be fruitful.  Once the subject and object, the knowing and the known, are seen as a unified event, what he called the transactional field, one can see that the valuation of resources depends on the acting subject.  What defines entrepreneurship is that the valuation takes place only through acting.  Thus, the acting agent, whose set of constraints becomes more defined by acting, is not powered by certain knowledge of the set of resources.  Such uncertainty, which may lead to anxiety, cannot be reduced to risk—where risk is the probability distribution of already known, routinized events.  Once the risk/uncertainty difference is admitted, it would be a short distance to propose an economic theory of entrepreneurship à la Dewey.


Key words:      risk vs. uncertainty, optimization, means-end in neoclassical theory, means-end in Marxian theory, belief, self-action, interaction, transaction.

JEL classification:    D0


The phenomenon of entrepreneurship is probably the most embarrassing chapter of neoclassical as well as Marxian economics.  The embarrassment is well warranted.  In fact, given their metaphysics, neoclassical and Marxian theories cannot provide a theory of entrepreneurship.

One of the major tenets of neoclassical, mainstream economics is the sharp separation of preferences (ends) and resources (means).  The separation is almost synonymous with economizing.  One has to subscribe to the means-end dichotomy if the problem is to select the cheapest mean to attain a given end or, what is the same thing, to use efficiently given means to attain the greatest end.  The dichotomy is also part of Marxian economics, where tools and materials are called "means of production" that can be objectively measured in terms of socially necessary labor-time and, hence, make it possible to measure the end, surplus value.  While the means-end dichotomy is useful to illuminate particular questions, it made it impossible to account for entrepreneurship.

            John Dewey questioned more than any philosopher the means-end dichotomy.  His critique of the dichotomy and its associated conceptual schemes stretches from his 1896 paper, "The Reflex Arc Concept in Psychology," to his 1949 book, Knowing and the Known, coauthored with Arthur Bentley.    I use here his view of means and end to show the way for a theory of entrepreneurship.  Section one shows why neoclassical, mainstream economics cannot account for entrepreneurship.  Section two shows the main reason why Marxian theory cannot account for entrepreneurship.  Section three exposes Dewey’s theory of action.  Section four provides a sketch towards a theory of entrepreneurship in light of Dewey’s theory.

1.  Neoclassical Theory of Choice

In standard neoclassical economic theory, the agent makes choices by optimizing his given end (the utility function, preferences, or the objective function) consisting of goods subject to given means (resources, budget, or constraint function).  The choice of how much to consume of each good would be complicated if we introduce different states of the world (such as health or weather), whether the objective function involves the utility of others (e.g., peer pressure, altruism, nationalism, and so on), or entails intertemporal utility dependency (the way neoclassical economists account for habit formation).  However, for the purpose here, we do not need to introduce such complications.  The purpose here is to capture optimization, the meaning of the standard notion of rationality.  Let us examine the simplest utility function, U, which consists of only two goods, X1 and X 2, with given prices, P1 and P2, that determines one’s budget constraint, I:

Max         U = U (X1, X2)

Subject to   I = P1X1+ P2X2

(given I ≥ 0; X1, X2 ≥ 0)

Here, we maximize U in light of the budget constraint and other technical conditions. 

From this simple set up, neoclassical theory constructs the laws of demand and supply.  Namely, if only P1 rises, the demander would substitute X1 with X 2, and the supplier would do the opposite.  Also, from this set up, neoclassical theory predicts how agents react to the expansion of income (I).  Although at different rates, the demand for both products usually rises with the rise of income.  In the case of necessities, however, agents would buy less when income rises if the good is inferior (such as public transportation).  While these laws, which Dewey and Bentley [1973] called interactional, have validity, their validity is not global.  Not all behavior can be explained in terms of response to stimuli such as the change of prices and income.   These laws assume, first, that the end is a given substance, second, that the set of means is a given substance.  What if neither is the case?

When the end or set of preferences is not given, violating the first assumption, an increase in price may attract more demand—even after controlling for the snob appeal or what economists call the Veblen effect [Leibenstein, 1950].  The rise of price, which is a barometer of difficulty, may prompt someone to inquire more intensely into the matter, hoping for greater experience.  If mountain climbing is an easier task than what it is, there would be much less demand for it [Loewenstein, 1999].  If reading Johann Wolfgang Goethe’s novels is as easy as reading Stephen King’s, less people would probably be reading Goethe today.  Just ask yourself:  How many people will be reading Stephen King’s novels in a couple centuries from now?  Probably none.  Agents do not seem to be exclusively interested in satisfying already given tastes.  They are also interested in satisfying vague, unclear tastes insofar as the satisfaction of theses tastes require effort, or production activity [Becker, 1965].  For instance, some agents expend enormous effort in sports, intellectual pursuits, sexual exploits, and even cooking in order to develop their vague tastes and unclear desires.  The demand to satisfy penumbral tastes seems to be a positive function of cost—a usually unnoticed violation of the law of demand.   Through the expenditure of effort, one comes to know and, hence, satisfy his preferences.

The violation of the first assumption characterizes what I call the "entrepreneurial consumer."   It is the violation of the second assumption, a given set of resources, that characterizes the "entrepreneurial manager"—equivalent to the traditional notion of entrepreneurship.  The entrepreneurial manager appears when the set of means or resources is not given.  In its extreme manifestation, the agent would imagine diverse worlds to the extent that he is insensitive to price change.  He would not worry about scarcity in the first place.  One could envision his own resources.  Thus, contrary to chapter one of modern college textbooks in economics, scarcity would not be the economic problem with which agents have to deal.  Here, one’s set of tastes would orient one’s cognition in one direction rather than the other.  If one, e.g., is interested in rustic restaurants, he would be able to see a train station as a potential rustic restaurant more than agents who do not have that taste.  Economists are famous for the statement that "there is more than one way to skin a cat," to signify the multiple-use of a resource.  Otherwise, the whole paradigm of neoclassical economics of efficient allocation would be dead on arrival.  However, the extent of usage of a resource in different deployments is far from clear.  A log can be used for firewood, a table, a chair, and so on.  Only an acting subject, a subject with a particular end in mind, can see a particular usage which, to his knowledge, no one has envisioned before.  Innovation is born.

Innovation, therefore, is about acting and conceiving.  Neoclassical economists in general cannot account, at least at first approximation, for innovation for an obvious reason.  Given the starting point of optimization, the subject and object are given substances.  Therefore, the subject cannot be innovative.  He can only rearrange material resources that are given and equally clear to everyone.  So, neoclassical theory, which is forced to reckon with innovation, introduces innovation as some kind of "manna" falling from heaven.  In neoclassical theories of growth, e.g., innovation is modeled as the result of a predictable mechanism along an iid process, Markov process, or some other statistical manner.  Neoclassical theory cannot, without ad hoc qualifications, account for innovation as an endogenous aspect of action as long as it conceives action as an interaction in Dewey and Bentley’s [1973] sense of the word.  Interaction takes the interacting parties as the same before and after the interaction.  It does not transform the acting parties.  It is simply a choice or a response undertaken by a pre-constituted subject (given set of preferences or ends) in light of the stimulus afforded by the pre-constituted object (given set of means). 

2.      Marx’s Concept of Abstract Labor

In neoclassical theory, the acting subject cannot be an entrepreneur because the set of objects are defined independent of him.  In Marx’s theory as well, the producer cannot be an entrepreneur because the set of objects, which Marx calls means of production, are defined independent of the producer.  To show this, let us examine the case of simple commodity exchange where each producer only employs himself—there is no capital hiring labor.

According to Marx’s labor theory of value, the value of the product is determined by the socially necessary time it takes to produce.  This includes the value of intermediate goods, the value of depreciation of machines and living labor—all evaluated in terms of labor time that is socially necessary.  Modern Marxist theorists have focused on the double meaning of "socially necessary."  It means, first, that the laborer should be exerting average social skill and, hence, the labor of higher skill labor should be compensated by a factor equal to the level of the skill.  Second, the amount of labor that counts as valuable is the one expended while so far the supply equals effective demand, i.e., demand determined at the "natural" value.  While these two aspects have been exhausted in the literature, Marx advanced another condition which received little attention.

Namely, for the labor value to meaningful, Marx assumed that labor power is "abstract" besides being "concrete" [Khalil, 1994].  For goods to be exchanged for the same thing, money, Marx thought that labor time must be of the same quality.  How can compare the labor time of a butcher and the labor time of a baker if they cannot be reduced to an abstract, homogeneous quality?  If the labor effort in each endeavor is only concrete, then one would be like counting apples and oranges.  There must be something abstract about labor process that allows us to say that one hour spent in baking is not differentiable from an hour spent in cutting carcasses.  Both labor activity, while concrete, are abstract and, hence, are equal in value.  To explain abstract labor, Marx [1976] discusses the labor process in chapter seven of Capital, Vol. 1.   The labor process involves man confronting nature, where man exerts physiological effort to extract goods from nature.  In the effort, he implies tools have to conceive the activity in the abstract before proceeding.  For Marx, the honeybee is "programmed" to produce honey and, hence, it labor cannot rise beyond the concerete.  However, human labor is conceived prior to execution—since humans are in command over nature with the help of technology, while other animals are not.

The basis upon which Marx justifies his concept of abstract labor is of great importance and has escaped the attention of many of his students.  The basis, the idea of abstract labor, is reasoned on the ground that conception of laboring precedes the action. In this sense, the action is merely instrumental, mediating between the conceived plan and nature.  Humans supposedly are different from animals in this regard:  Humans are in command of their environment because they conceive the act prior to acting, while animals are supposedly not.  To be in command, it means that humans see their labor as a tool external to them, which can be maniputlated, to contral the environment.  As such, it is not only the environment externalized, but also human labor.  The action does not transform the conception.  This is the case, for Marx, for the pre-capitalist mode of production, where the means of production are primitive and human conception was a product of action.  In modes of production characterized by primitive technology, labor was never abstract, which made it impossible for the rule of the law of value, as is the case in market-cordinated production activity.  In such modes of primitive technology, the human potential to be abstract is suppressed, which is the source for an alienation that differs from the one which inflicts the market and occasions commodity fetishism.  As discussed in detail elsewhere [see Khalil, 1992], the alienation occasioned by prmitive forces of production give rise to the worship of rulers and kings, what I call "rank fetishism," as a way for man to assert his power over nature.

So, for Marx, man stands clearly outside of nature—both natural environment and physiological body—and nature obeys.  Similar to neoclassical theory, Marx views resources (including his human labor) as given independently of experience.  Within this frame, there is little room for entrepreneurhsip.  The resources cannot help the agent to develop and conceive  resources in an innovative manner.

3.  Dewey’s theory of Action

In his famous article, "The Reflex Arc Concept in Psychology," which has been cited over a century mainly in the psychological literature, John Dewey offered a profound critique of interactional models.  To note, he came to call such models "interactional," as opposed to his alternative "transactional," fifty years later in his co-authored book, Knowing and the Known.

The object of Dewey’s critique is the reflex arc concept, what came to be known later as the stimulus-response (S-R) model or behaviorism.  Dewey re-examines the case of the child’s withdrawal of hand (response) upon feeling the burn of a candle (stimulus):

The ordinary interpretation would say the sensation of light is a stimulus to the grasping as a response, the burn resulting is a stimulus to withdrawing the hand as response and so on.  There is, of course, no doubt that is a rough practical way of representing the process.  But when we ask for its psychological adequacy, the case is quite different.  Upon analysis, we find that we begin not with a sensory stimulus, but with a sensori-motor co-ordination, the optical–ocular, and that in a certain sense it is the movement which is primary, and the sensation which is secondary, the movement of body, head and eye muscles determining the quality of what is experienced.  In other words, the real beginning is with the act of seeing, it is looking, and not a sensation of light.  The sensory quale gives the value of the act, just as the movement furnishes its mechanism and control, but both sensation and movement lie inside, not outside the act [Dewey, 1972, p. 97]. 

Put differently, for a stimulus to be a stimulus it has to be interpreted by the subject.  It does not exist independently, as something in-itself.  For an object to be sensed, the subject has to move and the sensory organ has to be positioned in a particular way.  What is wrong with the S-R model, for Dewey, is that it does not tell the whole story.  As a result, it presents the sensation (stimulus) and the movement (response) as external to the act:

[T]he reflex arc idea, as commonly employed, is defective in that it assumes sensory stimulus and motor response as distinct psychical existences, while in reality they are always inside a co-ordination and have their significance purely from the part played in maintaining or reconstituting the co-ordination; and (secondly) in assuming that the quale of experience which precedes the "motor" phase and that which succeeds it are two different states, instead of the last being always the first reconstituted, the motor phase coming in only for the sake of such mediation [Ibid., p. 99].

The S-R model presents a disjointed psychology, for Dewey.  It views the agent and the environment as independent of each other, connected by an arc, rather than seen as interdependent, constitutive part of a circle, what Dewey and Bentley called the "transactional" field over fifty years later.  In the S-R model, the actions of agents are conceived as a series of jerks to "pressures" imposed by external objects.  The S-R model is based on the subject/object dichotomy, which Dewey criticized more forcefully and directly in Knowing and the Known.  So, the S-R model resembles neoclassical theory—where optimization assumes that the agent’s means (stimulus) are independent of the end (response).  Neoclassical theory also fails to conceive the means as having a meaning accorded by the end itself.

In Experience and Nature, at the outset of chapter four, "Nature, Means and Knowledge," Dewey explicitly attacked the "formulation of classic political economy" for treating means (including human labor) as objects independent of the end or the acting subject.  In economic theory, labor is modeled as burdensome—expended only to support enjoyment.  In this regard, economic theory does not differ from the mythology of the Fall, where Adam and Eve were punished:

            No mythology is more familiar than that which tells how labor is due to trespass of man upon divine prerogatives, an act that brought curse upon the earth and woe to man.  … The tale is touching evidence that man finds it natural that nature should support his activities, and unnatural that the burden of continued and hard endeavor should be placed upon him.  Festivity is spontaneous; labor needs to be accounted for.  There is a long distance between the birth of the old legend and the formulation of classic political economy; but the doctrine of the latter that labor which is the source of value signifies cost, onerous sacrifice of present consummation to attainment of later good, expresses the same human attitude [Dewey, 1958, p. 121].

Dewey continues to note that laboring is "the source of all noting and recording of nature’s doings" [Ibid.].   Without labor, it would not be possible to respect the operation of events according to causality.  For Dewey, "[w]hile leisure is the mother of drama, sport and literary spell-binding, necessity is the mother of invention, discovery and consecutive reflection" [Ibid., pp. 121-122].   It is true that labor and its tools figure in work "not as fulfillments, realization, but in behalf of other things of which they are means and predictive signs."  But a tool is more than an instrument, "it is a thing in which a connection, a sequential bond of nature is embodied" [Ibid., p. 122].

            But why did theorists come to think of means as objects that exist "in-themselves" rather than constitutive of the existence of the subject?  What is the cause of the subject/object dichotomy that leads to conceiving resources as objects independent of action and, hence, stripped from meaning?  In The Quest for Certainty, Dewey [1960, p. 22] argues that the "quest for certitude has determined our basic metaphysics."  Man is after eternal forms that do not change:  "If a thing changes, its alteration is convincing evidence of its lack of true or complete Being" [Dewey, 1960, p. 19].  Practical action is needed to solve problems and, by definition, involves change.  In contrast, knowledge, as assumed by Greek philosophers, is after eternal Being that does not change.  Knowledge is seen as a reflection of nature and, hence, is elevated to a status higher than practical action.  So, the object of knowledge, ultimate Being, is fixed and given.  It cannot be altered by the knowing subject or by practical action.  So, the dichotomy between knowledge and action, which still lingers in modern philosophy, is the result of the quest for absolute certainty.  While knowledge has a "rational, necessary and unchanging form," practical action deals with change and, hence, it is contingent, a matter of probability, not of certainty [Ibid., p. 21]:

            [T]he main tradition of western culture has retained intact this framework of ideas.  Perfect certainty is what man wants.  It cannot be found by practical doing or making; these take effect in an uncertain future, and involve peril, the risk of misadventure, frustration and failure.  Knowledge, on the other hand, is thought to be concerned with a region of being which is fixed in itself.  Being eternal and unalterable, human knowing is not to make any difference in it. [Ibid., p. 21].

For western philosophy and culture, Dewey continues, knowledge does not depend on practical action. Thus, knowledge supposedly corresponds with what is real--and what is real is the "completely fixed and unchanging" [Ibid.].

            In Knowing and the Known, Dewey and Bentley [1973] emphatically rejected that the real is given, only waiting to be discovered by pure theorizing to produce knowledge.  They argued that knowledge is action and action is knowledge.  As such, knowledge is not a self-constitutive entity that reflects the real, which is supposed to be given and fixed.  To avoid the connotation associated with the term "knowledge," underpinned by the subject/object dichotomy, Dewey and Bentley proposed the terms "knowing" and "known."  These terms inter-define each other to emphasize that what is known depends on the active subject with his perspectives, questions, and ends.  As such, knowing is not a mirror that reflects the known. [1]   And the known is not a mount to be discovered.

            Dewey and Bentley [1973] called their metaphysics "transactional" to emphasize its difference from the standard, interactional models in science.  Interactional models, such as behaviorism and neoclassical economics, posit the subject and the object as given, and the theory supposedly shows the rules of their interaction.  Such interaction cannot transform the subject and object.  In contrast, Dewey and Bentley maintained that the subject and the object could only be defined in terms of each other within the transactional field.  Still, Dewey and Bentley did not expunge interactional models altogether for two reasons.  First, there are areas, such as the motion of bodies and friction, where the interactional models are valid.  In economics, arbitrage activity, behind the self-correction of markets in booms and busts, is nothing but agents involved in "interaction."  Such activity restores market equilibrium.  Second, for Dewey and Bentley, interactional models in science, stretching from Newton’s first and second laws of motion to Watson’s behaviorism in psychology, are improvements over "self-actional" models.  The models of self-action make the agent impervious to change in light of change of environment.  For example, despite the rise of P1, the consumer and the producer would continue with their activity.  They are, supposedly, operating according to an internal program, some kind of given entity inside their brain or being, that sometimes called "essence" "culture," or "habit."  In economics, the work of Thorstein Veblen and his followers, known as old institutional economics, comes closest to such self-actional models.  But full-scale self-actional models is expressed in Aristotelian physics, where bodies fall and air rises because of some kind of essence in their nature—so that iron would have greater acceleration while falling than cotton would have.  Galileo Galilee undermined the self-actional program and ushered the interactional model in theorizing.  According to Dewey and Bentley, the Galileo’s revolution, the interactional model, has been the hallmark of modern science.

While Dewey and Bentley reject self-actional models, and they find a role for interactional models, they warn that interactional models share a common element with the self-actional ones.  Namely, interactional models also presume that the actor is a given entity prior to interaction and even his constitution would not change as a result of choice.  According to the interactional models, the actor acts in response to stimulus, where the stimulus is metaphysically separate from the actor.  Such a view stems ultimately from the enduring subject/object dichotomy [see Bernstein, 1982], where the object is fully formed and the subject tries to remove the scales from its perspective to see the object in its supposed true existence. [2]

4.  What is Entrepreneurship?

Entrepreneurship is not simply about innovation, although innovation is the product of action per se.  To identify entrepreneurship with innovation may give the impression that entrepreneurship is much limited than is the case.  To understand entrepreneurship, we need to start with action per se, which is creative by definition [Joas, 1996].

Also, entrepreneurship is not simply about taking risks, although entrepreneurial activity involves what newspapers and the public call "risk taking."  To understand entrepreneurship, we need to investigate so-called "risk-taking."

One takes a risk, of miniscule value, even when one crosses a street in Manhattan, drives to work, or simply stays in bed at home and watch television.  There is always the lurking possibility of a hurricane, flash floods, earthquakes, or encounters with random criminals bursting into one’s bedroom while watching television.  One also takes a risk, of greater magnitude, when one converts all of his retirement savings into ownership of the stocks of a public company where he has no control.  Entrepreneurship, in the most general term, is more than risk-taking, even when it is of great magnitude.

As proposed here, entrepreneurship is about action when one’s ability is in doubt.  Ability can be in doubt even in the absence of risk.  But, of course, the doubt would be heightened in the presence of risk.  This heightening might be the reason behind the confusion of entrepreneurship with risk-taking.  Entrepreneurship as acting when in self-doubt is illustrated by the step a one-year old takes for the first time or the step a convalescing man takes after a long illness.  And even staying in bed, and watching television, can be entrepreneurial if one is in doubt about self-ability in case of how to handle a tornado or a fire.  For most people and for a multitude of occasions, however, walking and sleeping are routine acts and, hence, make choices as if self-ability is not in doubt.  In these cases, one reacts without reflection to stimuli.  Thus, one may apply in these cases the optimization calculus of standard theory of choice.

But when self-ability is in doubt, one cannot perform the optimization calculus even in probabilistic sense, using what economists call the "expected utility" calculus.  When one is in doubt, it is not that ability has a statistical likelihood of succeeding.  The person does not even know the probability distribution for the simple fact that he has not taken enough repetitive identical actions to form such a distribution—and even he may have not undertaken the act altogether.  The act could be unique.  On the ground of unique events, Frank Knight [1971] argued in his classic Ph.D. thesis that uncertainty cannot be treated as risk.

However, uncertainty is more than about unique events—otherwise it can be reduced to subjective uncertainty à la Leonard Savage as actually the case with Knightian uncertainty [Khalil, 1997].  Once one undertakes an action, one’s ability improves.  So, it is not even possible, in hindsight, to know if one could have done the act without acting, without experiencing one’s ability.  Experience, as Dewey stresses, is constitutive of existence.  The act itself reformulates ability in an irrevocable manner.  Still, in the face of a new challenge, or a task that one has not routinized yet, one would be in doubt again.  Once a toddler masters running, he wants to climb trees.  And once a child masters tree climbing, he wants to swim, and so on.

Self-doubt can develop into neurosis and psychosis, which the subject of psychology.  The agent can be gripped with fear to the point of being fixated by addictive substances, ruled by a course of action dictated by the "should" [Horney, 1950; see also Wachtel, 1983], or resort to leader worship [Fromm, 1994].  However, in the face of doubt, most people in most periods of history did not resort to such extreme measures.  They have usually adopted beliefs, or what I call in this context "paradigms" [Khalil, 2002], that abridge their self-doubt and allow them to act.

According to the standard theory of rationality, beliefs are rational when they are adopted in light of available information.  However, beliefs, and this is why I prefer the Kuhnian term "paradigms," are not falsifiable propositions.  If a twenty-years old man believes that he can be a great painter, one cannot refute it by appealing to the dismal statistics of likelihood of making it as a great painter, even in light of the past record of that young man.  One may state a counter statement, such as "you do not have the stuff that makes a great painter."  But such a statement is also a paradigm that cannot be falsified.

Even when the agent becomes an old man, and he spent all his life doing something else such as selling life insurance policies, this would not be a disproof of his original belief.  There are always extraneous, unavoidable circumstances that one can use as excuses and qualifications, the point highlighted by the well-known Duhem’s irrefutability thesis. [3]   However, the use of excuses, which are abundant, is not the reason of why belief is protected from falsification.  Rather, the old, life-insurance salesman can stipulate that he has the stuff to make him a great painter, but he lacked the sufficient belief to carry out the required discipline.  He would be full of "regret" for not being bold enough in one interview or an encounter.  While he believed in his talent, he was not free of doubt because belief is the daughter of doubt.  While belief is born by doubt, it does not eliminate doubt.

Belief, nonetheless, allows the agent to undertake entrepreneurial actions that one would not have taken.  The belief can range from the religious to the secular, from the personal to the communal, and so on.  The belief, to be sustained, has to be reinforced by a routinized activity with similarly minded people.  The belief provides the agent with a perspective or a focus that articulate resources in an innovative way.  This articulation would not have been experienced without the belief.  If one believes that he has the stuff to be a painter, he would become aware of his appreciation of color or dimension, for example.  It even allows him to see the world around him in a different way.  Here the attainable end, made possible by the belief, allows the agent to organize his senses and motor skills, to discover tacit or hidden means, and to realize the fruitation of the means.  Thus, means are not given as in the neoclassical or behaviorist sense, i.e., exist as self-enclosed things.  Means, as Dewey emphasized, have meanings.  They acquire their meaning in light of the context, the end empowered by the belief one comes to acquire.  So, tools including labor are not pre-constituted entities.  They are rather conceived and become tools in light of one’s end.

To view the budget constraint as dependent on the end violates the assumption that the budget constraint is scarce—an assumption needed to perform optimization. One cannot use things in the most efficient way if such things, their existence, depend on experiencing them via an end powered by a belief.  A tool cannot be used to maximize an end, such as utility or profit opportunity, if the end allows us to define the tool and its use to start with.  While neoclassical theory conceives the means as independent of end, their uncertainty allows the end to provide meanings to the means.

5.  Conclusion

Dewey’s transactional view resembles Kenneth Boulding’s [1956] emphasis on one’s image in the articulation of knowledge and Michael Polanyi’s [1958] stress on how knowledge is constructed upon the personal development of the knower.  These ideas resonate with the ecological theory of perception as advanced by James Gibson [1986; see Rogers, 1998].  The ecological theory regards perception as the activity of the whole body, where the brain is seen as part of the organism as it attempts to appropriate its environment [see also Damasio, 1994].  Such an embodied view of cognition can be traced back to the work of Kurt Goldstein [1995].  Goldstein examined the manner in which brain-injured soldiers cope with their environments.  The phenomenologists Maurice Merleau-Ponty [1994] and Alfred Schutz [1970, 1982] used Goldstein’s ideas to press on philosophers the centrality of the body and its needs to understand cognition, language, and behavior [see Kestenbaum, 1977; Rosenthal & Bourgeois, 1980; Lakoff & Johnson, 1999].

Further, Dewey’s transactional view of action resembles Thorstein Veblen’s [1898] critique of utilitarian economics.  Followers of Veblen, old institutional economists [e.g., Hodgson, 1993, 1998], point out how neoclassical economics presents the agent as an automaton that responds to stimulus, ignoring how habits make sense and organize the agent’s environment.  Likewise, Friedrich Hayek [1952], a leading exponent of Austrian economics, levels a critique of behaviorism that resonates well with Veblen’s 1898 essay and Dewey’s 1896 essay.  Hayek argues that the agent organizes the observed field according to tacit knowledge of which even the acting agent, not to mention a central economic planner, is ignorant. 

However, many Austrian economists, advancing a radical form of subjectivism, emphasize knowing at the expense of the known [passim Lavoie, 1990]. They usually stress linguistics over experience, as if the known is fully manipulated by the subject. However, as Dewey and modern critics of post modernism have warned, subjectivism may lead to solipsism.  Whether the subjectivism of some Austrians [e.g., Lachmann, 1976; Shackle, 1952, 1970] is solipsist is open to debate.  But, with the threat of solipsism, some Austrian economists [e.g., Kirzner, 1973, 1985] have, effectively returned unwittingly to the arms of neoclassical theory.

            While this essay is not on Austrian economics, one can appreciate Dewey’s transactional approach better as one learns from the debacle Austrian economics has faced in its treatment of entrepreneurship. The main fame of Austrian economics is that it can, while neoclassical theory cannot, account for entrepreneurship.  The subjectivists argued that opportunities are not objective, but rather a product of the imagination and cognitive apparatus of the purposeful agent.  In this view, purposefulness was the entry point to theorizing to the extent that purposefulness came to signify a mental capacity.  With such a hardy view of the acting subject, the object became fully dissolved, as if resources are totally the invention of the imagination.  In this light, entrepreneurship became synonymous with romantic views as if experience does not matter.

            According to Dewey/Bentley’s transactional field, the object is not constructed as one imagines.  Rather, the object places constraints on imagination.  The product of thought is the unity of the subject and object, not the dissolution of the object totally into the subject.  Innovation is not an arbitrary, whimsical activity for Dewey.  In "The Postulate of Immediate Empiricism," Dewey argues, "things are what they are experienced to be" [Dewey, 1977, p. 159].  And the way things are experienced depend greatly on subtle, non-conscious primal experience.  Such primal experience places a limit on how the agent understands encountered resources.

            To rectify a major failing in neoclassical theory, we need to understand that means carry meanings conferred by the acting subject, the entrepreneur.   Such a transactional view of action need not lead to subjectivism if we equally understand that the acting subject is empowered with a perspective or a set of ends that are informed by previous experiences.  Such a view connects action to the process of change, and allows us to view the evolution of institutions as the result of an endogenous process.  But this would be another story.


Becker, Gary.  "A Theory of the Allocation of Time."  Economic Journal, 1965, 75, pp. 493-517.

Bernstein, Richard J.  Beyond Objectivism and Relativism: Science, Hermeneutics and Praxis.  Philadelphia: University of Pennsylvania Press, 1983.

Boulding, Kenneth E.  The Image: Knowledge in Life and Society.  Ann Arbor: University of Michigan Press, 1956.

Damasio, Antonio R.  Descartes’ Error: Emotion, Reason, and the Human Brain.  New York: G.P. Putnam’s Sons, 1994.

Dewey, John.  "The Reflex Arc Concept in Psychology."  In The Early Works, 1882-1898, 5: 1895-1898.  Carbondale and Edwardsville: Southern Illinois University Press, and London and Amsterdam: Feffer & Simons, 1972, pp. 96-109.  (Originally published in Psychological Review, July 1896, 3, pp. 357-70.)

________.  "The Postulate of Immediate Empiricism."  In The Middle Works, 1899-1924, Volume 3: 1903-1906, ed. by Jo Ann Boydston, associate textual ed., Patricia R. Baysinger, with an intro. by Darnell Rucker.  Carbondale and Edwardsville: Southern Illinois University Press, and London and Amsterdam: Feffer & Simons, 1977, pp. 158-167.  (Originally published in Journal of Philosophy, Psychology and Scientific Method, July 1905, 2:15.)

________.  Experience and Nature.  New York: Dover Publications, (1929) 1958.

________.  The Quest for Certainty: A Study of the Relation of Knowledge and Action.  New York: G.P. Putnam’s Sons, (1929) 1960.

________ and Arthur F. Bentley.  Knowing and the Known.  In Rollo Handy and E.C. Harwood (eds.), Useful Procedures of Inquiry.  Great Barrington, MA: Behavioral Research Council,  (1949) 1973.

Fromm, Erich H.  Escape from Freedom.  New York: Henry Holt, (1941) 1994.

Gadamer, Hans-Georg.  Truth and Method, trans. and ed. By Garret Barden and John Cumming.  New York: Seabury Press, 1975.

Gibson, James J.  The Ecological Approach to Visual Perception.  Hillsdale, NJ:  L. Erlbaum, (1979) 1986.

Goldstein, Kurt.  The Organism: A Holistic Approach to Biology Derived from Pathological Data in Man, with a foreword by Oliver Sacks.  New York: Zone Books, (1939) 1995.

Hayek, Friedrich A.  The Sensory Order.  Chicago: University of Chicago Press, 1952.

Hodgson, Geoffrey M.  "Institutional Economics: Surveying the `Old' and the `New'."  Metroeconomica, February 1993, 44:1, pp. 1-28.

________.  "The Approach of Institutional Economics."  Journal of Economic Literature, March 1998, 36:1, pp. 166-192.

Horney, Karen.  Neurosis and Human Growth: The Struggle Toward Self-Realization.  New York: W.W. Norton, 1950.

Joas, Hans.  The Creativity of Action, trans. by Jeremy Gaines and Paul Keast.  Chicago: University of Chicago Press, 1996.

Kestenbaum, Victor.  The Phenomenological Sense of John Dewey: Habit and Meaning.  Atlantic Highlands, NJ: Humanities Press, 1977.

Khalil, Elias L.  "Nature and Abstract Labor in Marx."  Social Concept, June 1992, 6:2, pp. 91-117.

________.  "Neo-classical Economics and Neo-Darwinism: Clearing the Way for Historical Thinking."  In Ron Blackwell, Jaspal Chatha, and Edward J. Nell (eds.) Economics as Worldly Philosophy: Essays in Political and Historical Economics in Honour of Robert L. Heilbroner.  London: Macmillan, New York: St. Martin's Press, 1993, pp. 22-72.

________.  "Buridan's Ass, Uncertainty, Risk, and Self-Competition: A Theory of Entrepreneurship."  Kyklos, 1997, 50:2, pp. 147-163.

________.  "Symbolic Products: Prestige, Pride and Identity Goods."  Theory and Decision, August 2000, 49:1, pp. 53-77.

________.  "Two Kinds of Institutions: Paradigms vs. Conventions."  A working paper, 2002.

Kirzner, Israel M.  Competition and Entrepreneurship.  Chicago: University of Chicago Press, 1973.

________.  Discovery and the Capitalist Process.  Chicago: University of Chicago, Press, 1985.

Knight, Frank.  Risk, Uncertainty and Profit, intro. by George J. Stigler.  Chicago: University of Chicago Press, (1921) 1971.

Lachmann, Ludwig M.  "From Mises to Shackle: An Essay on Austrian Economics and the Kaleidic Society."  Journal of Economic Literature, March 1976, 14:1, pp. 54-62.

Lakoff, Geroge and Mark Johnson.  Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought.  New York: Basic Books, 1999.

Lavoie, Don (ed.).  Economics and hermeneutics.  London: Routledge, 1990.

Leibenstein, Harvey.  "Bandwagon, Snob and Veblen Effects in the Theory of Consumer Demand."  Quarterly Journal of Economics, May 1950, pp. 183-207.

Loewenstein, George.  "Because it is There: The Challenge of Mountaineering … for Utility Theory."  Kyklos, 1999, 52:3, pp. 315-344.

Marx, Karl.  Capital, vol. 1, intro. by Ernest Mandel.  Harmondsworth: Penguin; London: New Left Books, 1976.

Merleau-Ponty, Maurice.  Phenomenology of Perception, translated from the French by Colin Smith.  London: Routledge, (1945) 1994.

Rogers, W. Kim.  "On the Mode of Being of Living Beings and Their Environment: Preliminary Ideas for an Ecological Approach in Philosophy."  Analecta Husserliana, 1998, 52, pp. 531-547.

Rorty, Richard.  Philosophy and the Mirror of Nature.  Princeton, NJ: Princeton University Press, 1979.

Rosenthal, Sandra B. and Patrick L. Bourgeois.  Pragmatism and Phenomenology: A Philosophic Encounter.  Amsterdam: Grüner, 1980.

Polanyi, Michael.  Personal Knowledge: Towards a Post-critical Philosophy.  London: Routledge & Kegan Paul, 1958.

Schutz, Alfred.  On Phenomenology and Social Relations; Selected Writings, edited and with an introd. by Helmut R. Wagner.  Chicago: University of Chicago Press, 1970.

________.  Life Forms and Meaning Structure, translated, introduced, and annotated by Helmut R. Wagner.  London: Routledge & Kegan Paul, 1982.

Shackle, G.L.S.  Expectations in Economics.  Cambridge: Cambridge University Press, 1952.

________.  Expectations, Enterprise and Profit.  London: Allen and Unwin, 1970.

Veblen, Thorstein.  "Why is Economics not an Evolutionary Science?"  Quarterly Journal of Economics, 1898, 12:3, pp. 373-397.

Wachtel, Paul L.  The Poverty of Affluence: A Psychological Portrait of the American Way of Life.  New York: Free Press, 1983.

     *  Behavioral Sciences Research Council, a division of American Institute for Economic Research,, and Department of Economics, Vassar College,  The paper benefited from discussions with Frank Ryan, W. Kim Rogers, Michael Lawlor and the comments of Robert Mulligan, Mathew Forstater, and Tucker Hughes. The usual caveat applies.

[1] Although with some differences, Richard Rorty [1979] and Hans-Georg Gadamer [1975] also emphasized the unity of the subject and object.

[2] The subject/object dichotomy informs not only neoclassical economics and behaviorism, it is also underpins Darwinian evolutionary theory [Khalil, 1993]—which Dewey failed to note in his excited response to Darwin’s Origin of Species.  For natural selection theory, the subject (organism) is presented as passive, succeeding or failing in response to the object (the selection force) that is assumed to be self-constituted and independent of the subject.

[3] It is also known as the Duhem-Quine thesis.  It the idea that no single hypothesis can be falsified because, when we test we test the hyposthesis with other hypotheses and supporting conditions.  One can always blame that the offered refutation is not valid because one supporting condition was absent.