2. Language and terminology

I have already noted that D&P's aim is nothing less than to establish foundations for evolutionary economics and, further, that they see the micro meso macro framework as a ‘single unified framework’ that can encompass and provide coherence to the ‘loose analytic consensus’ that currently exists in the area (D&P 2). In order for foundational/unifying exercises of this level of ambition to succeed, it is clearly necessary that the foundations laid – whether these be epistemological, mathematical, ontological, or anything else – are secure enough to support whatever is to be built on them. And a key ingredient here, especially for foundations intended as the basis for theory development, is that they be well defined and internally consistent.

Much of what follows below is concerned with how D&P's account, laudable as it is in other respects, falls short in this regard. Before I get on to the details, it might be useful to identify the two main sources of the problem. The first is that D&P tend to use terms in ways that sometimes diverge significantly from their commonsense meanings. This is always a risky strategy, even when accompanied by a glossary and warnings that the exposition is conducted ‘in a new language of [their] own creation’ (D&P xiii), because it is often difficult for the reader to fully escape commonsense meanings and associations and thereby avoid being misled. The problem is then exacerbated, and this brings me to the second source of the problems in D&P's book, when accompanied by various inconsistencies, equivocations, and vaguenesses in the terminology used.

Since this last charge is a serious one, let me provide some examples. Consider D&P's use of the term ‘generic’, a key term in their analysis and which will also be required below. D&P use the term copiously throughout the book, in references to ‘integrated generic framework’, ‘generic economy’, ‘generic evolution’, ‘generic change’, ‘generic rules’, ‘generic choice’, ‘generic micro coordination’, ‘generic macro coordination’, ‘generic individualism’, ‘generic perspective’, ‘generic endogenous growth theory’, ‘generic policy’, and so on. It is clear that ‘generic’ is a guiding conception in their scheme, which then immediately begs the question of what it means. Mindful of D&P's warning that they are dealing in a new language of their own creation, let us begin with what they say in the glossary:

While this definition appears straightforward enough when taken in isolation, problems begin to compound when the reader tries to hold to it in navigating the book as a whole. In the first place, an immediate problem that arises is that it is hard to separate it from the definition of ‘meso’ given just a page further on:

The only apparent difference between the first definition and the first (key) line of the second is that whereas the former refers to rules and their carrier populations in the plural, the latter refers to a rule and its carrier population (what D&P call a ‘meso unit’) in the singular. What is to be made of this? Do D&P intend the difference to be significant? On the one hand, given the almost identical way in which the two definitions are phrased and the observation that ‘Meso is always generic’, the reader might be forgiven for concluding that D&P regard the two domains as one and the same thing, i.e. that the meso is the generic part of the economy. On the other hand, if the difference between the two definitions is regarded as significant, then the phrase ‘Meso is always generic’ might be read simply as saying that as the generic concerns all rules and their carrier populations, it naturally encompasses any particular meso unit. This seems closer to what D&P intend by the phrase, in my view − although, for reasons that will emerge below, I don't believe that even this captures their intention (which is to point to the generic quality of meso, a quality that is however not mentioned in their definition of ‘generic’).

A second and more serious problem with D&P's general definition of ‘generic’ is that it is inconsistent with how they use the term in other places. Consider the following definition, which appears on the very next page:

While according to D&P's general definition, the ‘generic’ explicitly comprises the domain of rules and their carrier populations, micro – and therefore micro generic – is explicitly restricted to the individual rule carrier as ‘the building block of meso analysis’ (D&P 27). D&P cannot have it both ways. Either the general definition of generic needs to be reformulated in way that avoids reference to carrier populations, or the definition of micro needs to be revised so as to include carrier populations as well as individual carriers.

It is probably unfair to focus exclusively on glossary definitions, so let us see what can be gleaned about the meaning of ‘generic’ from the text itself. The most revealing passages I could find run as follows:

They go on:

In these passages D&P appear to use ‘generic’ as an adjective to distinguish social phenomena – rules in particular – as emanating from, and residing in, the human mind, irrespective of whether this is in the context of micro, meso, or macro. This interpretation makes sense, and, as far as I can tell, is the one used most consistently throughout the book. It also avoids the two problems I started out with, and is therefore the one that I will try to hold on to in the remainder of what follows. However, if this really is what D&P mean by ‘generic’ then little hint of this is given in their glossary definition. Indeed, the definition given there obscures this interpretation because, as stated, it applies as much to populations of non-sentient biological (genetic) rule carriers that D&P mention elsewhere as it does to human (generic) rule carriers.

3. Generic analysis and the micro meso macro framework

Having said a bit about what D&P might mean by ‘generic’, we are now in a position to provide a brief description of what they call generic analysis and their micro meso macro framework. With respect to the former, D&P begin with the observation that the major source of economic growth at, at least in the current stage of human history, is the growth of knowledge and what they call the re-coordination of knowledge (by which they mean the fitting together of existing bits of knowledge in an efficacious way). Economic evolution is in turn an expression of the growth and re-coordination of knowledge, and generic analysis the analysis of this evolutionary process: ‘generic analysis because it can be decomposed into analysis of rules and populations of rule carriers. The concepts shall be the building blocks of our generic analysis of economic evolution in the modern market economy in which scarcities and constraints are resolved with new knowledge, and in which wealth is predominantly created from knowledge’ (D&P xii).

Perhaps the best way for a newcomer to get a handle on the micro meso macro framework is by considering the constituents of each of its three domains. Micro is the domain of ‘the elementary unit of economic evolution’, the individual ‘rule carrier’ who uses rules for ‘operations’ (transactions and transformations) and who originates, adopts, and retains novel generic rules. Micro analysis consists of the analysis of instances of what D&P call a ‘generic micro trajectory’, namely the process by which a micro unit acquires a new generic rule and thus changes its knowledge base (D&P 36–43). The full story is a complicated one, involving various categories and orders of rules (see below). But the essentials are that (1) new generic rules originate in the mind of individual agents; (2) those same agents decide whether to adopt those rules (on the basis of whether ‘they expect the new rule will contribute positively to their generic capabilities and operational outcomes’ (D&P 40), and (3) then retain those novel rules ‘for ongoing use’ (D&P 43).

It may sound curious to think of agents as rule-originators who then decide whether they are going to adopt and retain the rules concerned on the basis of what sounds very much like a weighing of their adoption costs and adoption benefits (D&P 41). After all, most of the social rules that structure the world we are born into are not ones that we can in any sense decide. It will therefore probably make things clearer for the moment to bear in mind that when D&P speak of rules ‘originated’ by agents in the present context, they are often referring to what would normally be called new technologies, and then to product innovations in particular. The agents that D&P have in mind, in other words, are often what would ordinarily be called entrepreneurs engaged in innovation. I come back to the issue of their use of the term ‘rule’ as a synonym for innovation, as well as all manner of other things, in Section 7 below.

The second ‘meso’ domain consists of rules and their carrier populations. The meso is where evolutionary action takes place, according to D&P, where new rules are adopted by populations. The structure of D&P's account of meso mirrors their account of micro. Again, the object at the center of analysis is a ‘rule trajectory’, in this case the meso rule trajectory that describes how a new rule originating from a single rule-carrier spreads via adoption by other rule-carriers. Here too the process is broken down into three stages, ‘of the origination of a rule as a discovery (invention), its adoption into a population of carriers as evolutionary dynamic, and its retention by that population as an (evolved) institution’. Thus:

Finally, the macro domain comprises meso units (populations), and the purpose of macro analysis is to understand how the meso units of which the macro structure is composed are coordinated, and how that structure may change as a consequence of changes at the meso level. The story involves a novel distinction between macro order (coordination at the deep level of rules) and macro equilibrium (coordination at the surface level of populations) (D&P 61–62), and takes the form of a generalized account of how coordination is achieved over what they call a macro trajectory. As in the case of micro and meso, this is presented as a three-stage process, in this case involving: (1) de-coordination of order through the emergence of new meso (i.e. the emergence of a new technology, opening up new possibilities, leading to the formation of new populations of ‘rule carriers’); (2) re-coordination of order through local and global population change; and (3) coordination of new order that is retained, embedded, and replicated.

That completes my sketch of the micro meso macro framework. I have of course not attempted to convey anything of the depth and intricacy of D&P's analysis, which is considerable. I have also left aside the final part of the book in which they apply their framework to questions of growth and development (Chapter 6) and what they call generic policy (Chapter 7). As elsewhere in the book, the analysis in these chapters is complex and D&P have many interesting things to say, e.g. that economic policymakers should concentrate on intervening in the generic domain rather than at the operational level, and then, for reasons they spell out, by focusing exclusively on some varieties of rules and not others (D&P 94–99).

4. Ontology

We are now almost in a position to consider where ontology fits into D&P's project. Before we do so, it is probably useful to say a little about what ontology is. The term derives from the Greek, where ‘onto’ means ‘being’ and ‘logos’ can be understood as ‘study’, ‘science’, or even ‘theory’. Ontology is thus the study or science of being. The word ‘being’ has two senses, however, referring both to what it is to be or exist, and to something that is or exists, an entity or thing. On this line of thought, ontology extends to the study of both (1) what it is to be or to exist and (2) the study of entities or things that exist. Following Lawson (Reference Lawson2004), I will call the former philosophic ontology and the latter scientific ontology.

The scope of ontology as I have defined it is very wide, and any particular ontological project must adopt a particular focus of study, that is address some entities or things that are regarded as significant and worth investigating from an ontological point of view. The ontological theorizing conducted by D&P falls into the category of social ontology, that part of ontology concerned with the social realm. By the social realm I mean the domain of phenomena whose existence depends at least in part on the existence of human actors. Social phenomena thus include social rules and practices, technological objects, institutions, nations and economies, and even social and philosophical reasoning itself. Social ontology also divides into its philosophic and scientific, on the same lines as described above.

While social ontology is well established as an independent branch of knowledge in its own right (e.g. Searle, Reference Searle1995), many who pursue it do so in support of other ends. A common motivation here is that ‘doing’ social ontology forces researchers to be clear about the nature of the entities that are (or might be) the subject of their investigations, and may provide a basis for deciding which of their aspects to focus on and how to do so. The different ways in which research may be informed as a result may be grouped into three broad categories. In the first place, ontological investigations may provide the basic categories of existences that form the basis (or more grandly ‘the foundations’) of abstract theory development. This is clearly what D&P have in mind with respect to their fundamental ontology:

The second way in which ontology may inform research is by providing conceptualisations of particular entities, e.g. social rules, technological objects, money, the corporate form, etc., and by unpacking their structure, properties, and powers. As we will see below, most of D&P's ontological theorizing is actually of this second kind, in the form of an extensive analysis of generic rules. Third, and finally, ontology may have a significant role to play with respect to resolving methodological issues, e.g. by providing a basis for clarifying and resolving differences between rival theories and for exploring possible connections and complementarities between them, and by providing a basis for making informed decisions about the suitability of different methods of analysis in the light of the nature of the material they are being applied to (given that not all methods are equally well-suited to the analysis of different kinds of phenomena).

5. The ontology of evolutionary realism

We are now in a position to consider the ontological foundations of the theory just described, that is D&P's ‘ontology of evolutionary realism’ and what they have to say about social rules. D&P (3) begin by describing their fundamental ontology, which they present as being ‘composed of three empirical axioms’:

These axioms all fall within the category of philosophic ontology, as they refer to what it is for things to exist. D&P provide a brief elaboration of what each of these axioms mean, reversing the order in which they are stated above:

They then sum up their ontology of evolutionary realism as follows:

There are clearly some interesting ideas here. Remarkably, however, and save for a brief recapitulation and a few stray observations on the following page, this is pretty much all that they have to say about it. And as they do not attempt to unpack their axioms much beyond the short descriptions quoted above, their account accordingly raises considerably more questions than it provides answers. In what sense are the axioms empirical, if by empirical we mean derived from, or relying on, observation, experiment, or experience? In what sense are all entities supposed to be processes? Why should the economy being located in time and that all entities have a temporal existence necessarily be any indication that all entities are processes? What exactly is meant by ‘association’ (even if between ideas) and what is the connection with the existence of structure? What exactly is a ‘matrix of associations’? What do ‘associations’ consist of? And so on.

In short, D&P's fundamental ontology is breathtaking in its brevity – not least in view of the fanfare that precedes it – and it is therefore difficult to engage with it without a great deal of speculation about what they might have in mind. In one sense this doesn't matter too much, as D&P barely mention their three axioms of evolutionary realism again after introducing them. Very little seems to turn on this, moreover, as it quickly becomes clear that the real action from an ontological point of view lies in their analysis of social rules. However, as Axiom 1 does bear on D&P's treatment of rules, I shall say a little bit about it here.

Axiom 1 is the claim that everything that exists in an evolutionary economy is an actualization of some idea(s). This claim recurs in the definition of ‘bimodality’, the term that D&P use to refer the property expressed by axiom 1:

The problem is that D&P cannot literally mean what they say in this definition and in Axiom 1, unless of course the ideas they refer to include those of a superhuman intelligent designer (presumably not something that proponents of an evolutionary approach would want to argue). The reason for this is that there are all manner of things that exist in an evolutionary economy that are presumably not the matter-energy manifestation of ideas – or at least not human ideas – such as the human form itself, other fauna, flora, and the minerals and other non-biological endowments of nature used in production. There is also the problem that, however much they may bear the imprint of ideas, many social phenomena are unintended consequences of the activities of human actors, and therefore not the actualisation of a pre-existing idea, e.g. the market economy itself, unemployment, and the credit crunch.

The reader is therefore driven to conclude that the existences that D&P have in mind are restricted to ‘matter-energy manifestations’ that in some way bear the imprint of human ideas. And, indeed, there are hints in places that this may be D&P's position too. For example, while their Axiom 1 and definition of bimodal clearly refer to ‘all’ existences and are not in any way restricted to social phenomena, when they come to define ‘idea’ they refer specifically to what they call ‘generic existence’:

If the reference to ‘generic’ here is indeed intended to denote rules emanating from, and carried in, the human mind, then D&P would appear to be referring to existences that are social in the sense of depending on the ideas and activities of human actors. Some independent support for this interpretation is given by D&P's distinction between ‘genetic’ rules and ‘generic’ rules:

However, if the interpretation I have been urging is correct, then various adjustments have to be made to other parts of D&P's scheme. To give an indication of what is involved, consider what they have to say about rules. According to what appears to be intended as a general definition offered in the glossary, rules are ‘deductive procedure for operations’, where ‘operation’ refers to the transaction and transformation activities of human actors:

It is not clear why rules need be especially deductive procedures for operations (do D&P mean normative?). But leaving that aside, and if my suggested interpretation of generic is correct, then D&P's definition of the term ‘rule’ quoted above is not a general definition after all – unless, and as seems unlikely, ‘genetic rules’ are also to be regarded as ‘deductive procedures for operations’. A provisional solution to this problem may be to interpret the above definition of a ‘rule’, not as the general definition as it is presented in the glossary, but as a definition of a ‘generic’ (i.e. social) rule. We would then have:

Note that this definition is consistent with the description of generic rules in the passage quoted above distinguishing generic from genetic rules.

The problem raised above in connection with D&P's claim that all existences are the expression of some idea could then be dissolved by recasting Axiom 1 in a way that makes clear that it is referring not to all existences but only to those that flow – even if only indirectly – from human ideas and activities. As far as I can tell, an adjustment on these lines would not disturb their theory in any significant way. But D&P would be forced to pull back from their claim that their ontology of evolutionary realism ‘applies generally to all types of evolution, whether biological, immunological, . . .’ (D&P 4). If the ontology of evolutionary realism is restricted specifically to changes in generic rules, then it will have little to say about forms of evolution that involve changes in genetic rules.

6. Generic rules

I have argued that D&P's general philosophic ontology is highly underdeveloped, and that, save perhaps for the way that it informs the notion of all existences being matter-energy manifestation of ideas, it seems to have little bearing on how the micro meso macro framework was developed. Things are very different with respect to the (scientific ontological) development of social rules. D&P go into considerable detail here, and in a way that is fundamental to how they understand and elaborate the realms of micro, meso and macro.

We have seen that D&P define ‘rules’ as ‘[deductive] procedures for operations’ and I have suggested that this definition is best interpreted as capturing what they generally refer to as ‘generic rules’. One of the difficulties with D&P's account, however − and this will already have been apparent from some of what I have reported above − is that they use the term ‘rule’ in a way that stretches well beyond what would ordinarily be thought of as procedures for operations. I shall consider some of these other meanings in the next section, and use the present section to consider some of the specifics of generic rules as procedures for operations. D&P's (8) taxonomy of generic rules, reproduced in the table below, is a good place to begin.

The different categories of generic rules identified in this table are described as follows. Beginning from the left-hand side of the table, we have:

where:

Subject rules, according to D&P (7), thus ‘relate to an individual agent and define the internal rules of the agent's mind and the outwardly directed rules of the agent's behaviors’ (D&P 7).

On the right-hand side of the table we have:

where:

Object rules are ‘the class of rules for organizing things, including other people as rule carriers, that is, other subjects as objects’ (D&P 7).

Finally, in addition to the [CBST] taxonomy, D&P also distinguish three ‘orders’ of rules:

The description of 0th and 2nd order rules suggests that they may not be procedures for operations (it is only in the case of 1st order rules that D&P describe them specifically as such), in which case they would not be covered by the definition of generic rules. I shall not enter into this issue here and will concentrate mainly on the CBST taxonomy in the remainder of this section. But it is probably worth noting D&P regard 0th order rules as ‘the premises or axioms of an economic system embedded in social, cultural, political and legal rules . . . [concerning] such things as the role of the state in creating and implementing systemic rules, such as the rule of law, competition, monetary policy, property rights and suchlike’ (D&P 9). 2nd order rules, in contrast, are described as ‘a mechanism for changing [1st order] rules’ (D&P 10).

The categories listed above are of course only the skeleton of D&P's account, which they flesh out in some detail over the course of the book. I won't attempt to go into the many strengths of their analysis, save to draw attention to one of its most important features from an ontological point of view: their broad conception of generic rules as something that informs or guides what people do rather than being what they do, and the associated distinction between procedures and operations. This conception is crucial to their project because it places rules at the centre of the analysis as the relatively enduring structural component of society, but which may nevertheless be transformed or augmented through changes in human activities as the system evolves. On this front their account is well-aligned with other recent work, such as Feldman and Pentland (Reference Feldman and Pentland2003) on the distinction between the performative and ostensive aspects of routines in organization theory, and Lawson (Reference Lawson2003: 110–140) on the distinction between social practices as interactors and social rules as replicators in evolutionary economic theory.

Various problems remain however. In the first place, there are the definitional issues that I have alluded to in a number of places above. We have already seen that the definition of ‘rule’ provided by D&P is not a general definition because it excludes what they call genetic rules. Then we found that some rules (0th and 2nd order rules) may not be procedures for operations at all and therefore something distinct from what is defined in their glossary definition of generic rules. Added to all this is that reserving the term ‘social’ for the first of the two forms of object rules may be misleading since all of the generic rules referred to by D&P are social. These are tricky issues and I cannot claim to have spent the time required to sort them out. But a possible way forward might be to use the term ‘social rules’ or ‘generic rules’ for all rules that depend upon the existence of human agency (thereby excluding genetic rules) and then to treat what D&P currently call generic rules, i.e. procedures for operations, as well as 0th and 2nd order rules, and perhaps other categories of rules, as subsets of social rules. What D&P currently call social rules, if the category is still needed, might then be called institutional rules, or some such.

A second problem is that the subject rule/object rule dichotomy is a false one to the extent that it suggests that the rules that organize social organizations or technology (‘object rules’) and those governing the activities of individual actors (‘subject rules’) are different types of rules. In many cases, in my view, the opposite is true, namely that subjects and objects are governed by the same rules. Take the rules of email etiquette, for example, which serve both as what D&P call behavioral rules in governing the behavior of individuals engaged in emailing, and as what they call social rules in helping constitute the institution of email communication. Similarly, the constitutive rules that underpin the collectively accepted identity of a tool such as a Phillips screwdriver at the social level at the same time inform how individual users use such screwdrivers. The same may even be true in cases of what D&P call cognitive rules. For example, the South African 2007 Rugby World Cup winning team employed a sports scientist who introduced special exercises to improve the vision acuity, peripheral vision, and handspeed of individual players aimed at developing their ability to intercept passes made by opposing teams. The enhanced cognitive skills of the players then led, not only to their becoming better players individually, but also to successful new styles and strategies of play at the (social) team level. In short, the way that D&P cut up rules gives the impression that ‘subject rules’ are always distinct from ‘object rules’. I suggest instead that it is very often the same rules that are concurrently governing subjects and objects.

Note that if this last point is correct, it may also undermine what D&P (22–23) describe as their ‘generic methodological individualism’, exploiting ‘the aggregative relationship between micro and meso, such that the sum of carriers of a rule forms a meso unit (compose of a carrier population)’ (D&P 22). For while it may sometimes be useful to think of the meso as some kind of aggregation of rules originally emanating from the micro, it is surely also necessary to allow that existing rules relating to larger social objects, institutions, and so on, shape individual agents and how they behave at the micro level. And if so, generic methodological individualism would suffer from very much the same kind of deficiency as the more familiar ‘operant’ (D&P 22) version of methodological individualism associated with mainstream economics: i.e. just as there are no preconstituted ‘atomistic’ agents that exist independently of society and that can then be treated as the ‘building blocks’ of larger social phenomena, there may be no such thing as purely preconstituted rule carriers that can be analyzed independently of their location in society. A good example here is the digital camera: while its development certainly owed much to ideas emerging at the level of individual agents, it was also greatly affected by the social rules and routines associated with traditional photography (Runde et al., Reference Runde, Jones, Munir and Nikolychuk2009).

The final point I want to mention here is that there are difficulties with D&P's use of the term rule to refer to ‘ideas’ in the sense of an innovation or product. I come back to this issue in more detail in the next section, but the particular problem I want to note here is the idea that rules are the object of conscious, even maximizing, choices:

I do of course not want to deny that producers’ decisions about whether or not to develop and bring to market a product idea, or consumers’ decisions about whether or not to purchase a new product, are often made on a calculative basis of the kind described by D&P. However, and leaving aside whether it is helpful to describe people as thinking about whether or not to adopt a new rule in such cases (more on this below), it would be severely misleading for D&P's notion of generic choice to be extended to rules in general. While we may sometimes have occasion to reflect on a set of rules and decide to adopt them, in most cases the rules that affect us and which we live by are not the sorts of things that it is open to us to accept or reject (e.g. the rules of language, chess, general rules relating to the rights and responsibilities of social positions, etc.). Further, it seems to me that innovations tend to begin with the transformation of existing objects or practices (and sometimes the emergence of new ones) rather than social rules, and that is only when these catch on and spread that they begin to develop normative counterparts that we might want to call social rules. Far from being the objects of conscious choice, moreover, agents generally learn to act in ways appropriate to those rules in a subliminal way, perhaps even without their ever having internalized the rules concerned in even a subconscious way (Searle, Reference Searle1995: 137–147).

7. Casting the net too wide?

It will have emerged by now that despite the relatively narrow definition of ‘rule’ that D&P provide in their glossary, they go on to use the term to designate a remarkably broad range of things. These things include:

• • conceptions, schemata, attitudes and so on that exist in the mind, such as ideas, organizational principles (D&P 7), and preferences (D&P 75);

• • physical objects, such as ‘physical capital and commodities that embody technical rules’ (D&P 7); and

• • larger social objects like markets, ‘yet another generic rule’ (D&P 53).

While it may be argued that at least some members of the first of these three categories might qualify as ‘procedures for operations’, it is difficult to see how this could be true of physical or larger social objects. Perhaps all that D&P mean to claim is that rules are involved in organizing such objects. For example, in the sentence immediately following the one in which they say that ‘technical’ rules are ‘broadly defined by technology and include physical capital and commodities that embody technical rules’, they observe that ‘[t]hese are the rules for organizing physical things’ (D&P 7). But as so often elsewhere in the book, it is hard to be sure. On the one hand, they seem pretty clear that physical capital and commodities that embody technical rules are rules in their own right; on the other hand, it is hard to see how material objects like physical capital and commodities per se organize other physical things.

The root of the difficulty lies in D&P's use of the term ‘rule’ as the ‘analytical equivalent’ of ‘idea’. Consider the following statement:

This sentence starkly illustrates how D&P conflate ‘idea’ and ‘rule’, when the reader would have been far better served simply by being told about a match between supply and demand for a novel product (most innovators are after all not in the business of coming up with new procedures for operations in the first instance, and neither, for the most part, are procedures for operations traded on markets). What D&P mean when they say ‘rule’ in phrases like the ‘pop-up toaster rule’ (D&P 55–56), ‘the meso rule of electricity, synthetic materials and the internet’ (D&P 69), ‘the “internal combustion engine” rule’ (D&P 74) and so on, is what would ordinarily be called a ‘type’, i.e. taxonomic grouping of certain kinds of objects. The tokens of these types may themselves be (sub)-types – as in ‘variants on the “internal combustion engine” rule of two-stroke, four-stroke, diesel, rotary, multivalve, etc. . . .’ (D&P 74) – or their individual ‘matter-energy actualizations’. Clearly, it is generally only instances of the latter than can be traded on markets and adopted by users.

Note that I am not denying the importance of rules in constituting human artifacts and in governing our interaction with them. Indeed, as Phil Faulkner and I have argued elsewhere, the very identity of the artifacts that surround us depends on the functions that we collectively assign to those artifacts, and where those assignments are much like constitutive rules (Faulkner and Runde, Reference Faulkner and Runde2009). But that is not the same thing as saying that human artifacts are rules, or even that typifications are rules. In my view D&P's remarks about physical and larger social objects being rules should be rejected. In the case of physical objects, I would argue, it would be far better to say that common conceptions of types of objects, reflecting what agents know about their structure, functions, affordances, etc., condition agents’ interactions with those objects. And in the case of larger social objects, it would be far better to say that agents’ knowledge of – or at least capacity to behave in accordance with (Searle, Reference Searle1995) − the rules associated with them is what conditions their behavior. But considerably more work needs to be done to show how such knowledge translates into procedures for operations that inform the transaction and transformation of activities in which those objects are implicated.

No doubt D&P were striving for parsimony in their terminology and, in light of their view of social rules as the locus of evolutionary economic change, were driven to include as many things as possible under that heading. And no doubt there are occasions on which it is in the interests of an analytical exercise to use terms in ways that cast a wider net than they would in their ordinary language usage. But this, in my view, is not one of them. D&P would have been on far safer ground had they stuck rigorously to a more conventional conception of rules, and presented what they call ‘ideas’ as generally accepted ‘types’ along with an account of how these may condition how people interact with the physical or social objects concerned. Proceeding in this way would have facilitated a far more natural and accessible account in which (1) the product ideas of innovators could be called product ideas, (2) the new products that flow from these ideas could be called new products, and where (3) the transformed and newly emergent practices that flow from the adoption of these new products, when they become institutionalized, develop normative counterparts that are reproduced in and through those practices, and which we might want to call rules.