:: Article

Darwinian Creativity, Memetics and Some

Interview by Richard Marshall.

What philosophers add to empirical knowledge is conceptual clarity and normativity. That is particularly helpful if there is persistent disagreement over a long time without any progress in a debate that involves matters of facts. Such a debate usually involves misunderstandings that the philosopher can help dissolving – by focusing on the concepts involved. But analyzing concepts is not the only role of philosophy (understood as distinct from the sciences). Philosophers also ask normative questions. One normative claim I used in my research on analogies between evolution, creativity and culture is that one should not repeat trivialities.

Memetics reduces again either to something trivial – no big news that there are cultural items that spread, right – or to something false: that they make copies of themselves and that they spread because of their properties. The last point, an explanatory analogy, is often taken to also imply that memes spread independently of the beliefs and interests of human beings, which is also wrong…

There can also be an individual being that is the perfect example for a human being (regarding the properties that belong to the descriptive nature) but is not born to humans. Imagine that it came into existence in a different way, e.g. via a Frankenstein-like synthesis out of chemical material. Conceptually, this is possible. Many people, we assumed, would regard such an individual being as a person, but not as a human. So, where does that leave us?

Maria Kronfeldner works in the Philosophy of the Life Sciences and the Philosophy of the Social Sciences. She has published widely in this area, and has been awarded for some of her publications (The Karl Popper Essay Prize of the British Society for the Philosophy of Science; The Philosophical Quarterly International Essay Prize). Here she discusses Darwin and Lamarck, whether cultural change is Lamarckian, about the role of philosophy in science, the role of separation and integration in discussing culture and evolution, memetics, human nature and essentialism, the so-called ‘Interactionist consensus’ and genetic determinism, Collingwood and causation, and the issues of being interdisciplinary in university departments. This one plays for keeps…

3:AM: What made you become a philosopher?

Maria Kronfeldner: I was attracted to philosophy because it is the most minimalistic art I can think of. To express ideas, you use only the most minimal, the most reduced resources: no body (as in theatre), no figures (as in pictorial art), no voice or sound (as in music), no story (as in literature) – just thoughts. They are ordered, ideally crystal-clear and sharp, but they are just thoughts. There is beauty in this simplicity and at the same time a critical edge. It can elevate, it can hurt, as life.

3:AM: You work at the intersection of different disciplines – philosophy, anthropology, biological science. You’ve asked about whether culture is Darwinian from various angles – meme theories, synthesis in science, whether Lamarck is a viable alternative and so forth. Let’s start with your reflections on anthropology’s emergence from biology and the work of Alfred L. Kroeber who was following the work of the better known (to me at any rate!) Franz Boas at the start of the last century. So what are the boundaries between anthropology and the sciences looking like, and has it managed to establish a notion of ‘culture’ independent of heredity?

MK: Cultural anthropology has indeed established for itself some autonomy, after all it became a well-respected, separate academic field. Kroeber’s notion of cultural determinism was important for this (but certainly not the only important causal factor). He claimed that culture is explained with reference to previous culture – as cells are explained by previous cells. With this analogy (between culture and cells), he did not aim to bring culture and biology closer together; to the contrary, the aim was to distance the two by conceptualizing culture as an analogous, but autonomous kind of heredity, irreducible to cells, their heredity and the bodies made out of them. In order to create that distance between culture and biological heredity, Kroeber joined the anti-Lamarckian bandwagon that August Weismann was leading. Within a Lamarckian frame, culture and biology were interacting so much that the two were coupled, always changing together. That was a problem, certainly, for establishing autonomy.

3:AM: In biology, evolution is standardly Darwinian. You ask whether evolution in culture is not Darwinian but Lamarckian don’t you. So first, could you say what the difference is between the two approaches to evolution?

MK: First of all, Darwin was a Lamarckian in the sense that mattered to Kroeber: Darwin and Lamarck believed in the inheritance of acquired characteristics. Yet, they disagreed on how adaptation works. For Lamarck, ‘acquired’ simultaneously meant ‘directed’, i.e. caused in the direction of adaptation. Directedness in that context means that organisms do not develop variations in all kinds of directions of change. For Lamarck, the environment directs, i.e. determines, an adaptive change in organisms during the life of these organisms. Darwin believed, in contrast to Lamarck, that at least some changes in organisms are undirected. Hence if variation arises in all kinds of directions, a sorting process becomes necessary for evolution to take place. It follows that a Darwinian pattern of evolutionary change necessarily relies on a sorting process for there to be adaptive change – be this natural selection, sexual selection or what is now called drift, a kind of random sorting. Lamarck certainly saw that there is differential reproduction: that some organisms of a kind die earlier than others and reproduce less than others. But that played no role in his theory. In his picture, the complete population will sooner or later change in the right direction of adaptation, simply by the direction given by the environment.

If you apply these issues to cultural change, then there are a couple of questions that need to be distinguished. First of all, is culture Darwinian in the sense that it necessarily involves a sorting process of items that vary in all kinds of directions? Many cultural processes (e.g. the spread of artifacts or ideas) work that way; they are ‘variational’ or ‘populational’ (two terms currently in use for a decisively Darwinian pattern of change). Other cases of culture might not be variational. It’s an empirical question and distinct from two further questions. One is: is variation in culture as undirected as in nature? I tried to show in my work on creativity and Darwinism that this is not the case since our mind does direct the novelty produced by the goals involved. Third, you can also apply the concept of Lamarckian inheritance (the aspect Darwin and Lamarck shared) to culture and contrast it with a non-Lamarckian concept of inheritance.

And this is what I addressed in my paper on whether cultural evolution is Lamarckian. So, the question in that paper was: is culture a case of inheritance of acquired characteristics and in that sense Lamarckian? In one sense it certainly is. Culture is often defined as ‘non-innate’, but that is just another word for ‘acquired’. You do not get culture from your parents via sexual reproduction; you get it via social communication with all kinds of people, which can nonetheless be understood as a form of inheritance (as Kroeber did). But to claim on that basis that culture (i.e. cultural inheritance) is Lamarckian is a trivial claim: you define culture as acquired (rather than innate) and then you say that culture is inheritance of acquired characteristics. Such a claim is trivial since it is analytically true: its truth follows from the assumed definition of ‘culture’ (as acquired), as it follows from the definition of ‘bachelor’ that a bachelor is an unmarried man.

3:AM: Can you say in more detail why you conclude that cultural evolution in problem solving leads to what you call ‘directed variation’ which is Lamarckian?

MK: As said, that issue concerns the aspect where Lamarck and Darwin actually differed: whether variation arises in an undirected or directed way. Thus, the question is: is our mind creative (i.e. producing novel ideas) in a Darwinian (i.e. undirected) manner. I tried to develop a precise as possible formulation of what undirected versus directed variation could possible mean if we understand it as a general, formal principle (rather than a biological one). I finally ended up with three meanings of undirected variation. This precision was necessary since it seemed that in the debate about this, which was going on since a while, those involved were simply talking past each other. As before with the claim about whether inheritance is Lamarckian or Darwinian, I tried to show that with respect to one meaning of ‘undirected’ the claim is simply trivial: repeating what is included in a narrow definition of creativity (as requiring some spontaneity rather than foresight, which can be equated with a perfect directedness).

With respect to the other two meanings of ‘undirected’ that I distinguished, I argued that the claim is simply wrong. We are not randomly (one meaning) producing ideas. But since nature is not doing that either that was the easy case, a misguided meaning of undirectedness from the start. The most important meaning is the following: undirected means that the probability of the occurrence of new ideas is uninfluenced by the environment in which it occurs. This is the meaning of undirectedness that is so crucial for biological evolution being Darwinian, given current consensus. And with respect to that meaning creativity fails to be undirected since our minds (the environment in which selection of ideas occurs) is influencing the probability of the occurrence of an idea, at least in most cases that we count as creative. Previous knowledge makes later knowledge more or less likely.

3:AM: Is this a good illustration of the role of philosophy in science – it’s not about facts but about interpretation of the facts?

MK: I hope so. By distinguishing these different meanings of ‘undirected variation’, I tried to clarify an issue that also involves matters of fact, namely about novelty in nature and novelty in culture. What philosophers add to empirical knowledge is conceptual clarity and normativity. That is particularly helpful if there is persistent disagreement over a long time without any progress in a debate that involves matters of facts. Such a debate usually involves misunderstandings that the philosopher can help dissolving – by focusing on the concepts involved. But analyzing concepts is not the only role of philosophy (understood as distinct from the sciences). Philosophers also ask normative questions. One normative claim I used in my research on analogies between evolution, creativity and culture is that one should not repeat trivialities. This norm is part of argumentation theory: good arguments should not only be valid (having the correct truth-preserving deductive form) and sound (also having true premises); they should also be informative: they should be neither trivial (analytically true) nor irrelevant or redundant, i.e. repeating what is known since a while, but wrapped in a fancy new language, e.g. a Darwinian language.

3:AM: You argue that discussions about culture and evolution exhibit a bias towards synthesis, even though history shows that isolation also works. Can you unpack for us what the issue is here and what’s at stake?

MK: The informativeness of a knowledge claim is an epistemic value. Likewise the unity of knowledge is an epistemic value. We aim at a science that produces knowledge that is consistent. Consistency can refer to the parts of the theories internal to a discipline or field or to the relationships between theories of different disciplines or fields. Thus, we aim at synthesis (i.e. the unification) of knowledge. But synthesis procedurally viewed requires prior analysis (i.e. isolation, separation, choose your term). And this is often forgotten, e.g. when evolutionary psychologists criticize what they call the ‘standard social science model’, which assumes that social sciences can ignore claims about human nature and evolution for what they aim to explain. What is forgotten is that ‘isolation’ (e.g. of the social sciences) can also – as I try to show with the case of Kroeber and his defense of the concept of culture – be quite productive, for a while at least. So, it depends on the context, whether integration or separation (each taken as a research strategy) is more fruitful in order to arrive at new, i.e. at more knowledge. Given that separation as well as integration can be fruitful, I argue that they should be treated as equally valuable. But they are often not. This is what I call ‘synthesis bias’, an epistemically unjustified bias towards integration rather than separation.

3:AM: Memetics is supposed to explain cultural novelty and change analogous to genetics in biology. Can you first sketch out the theory of memes?

MK: Actually, it is hard to do, if you try to be precise. But a few pointers should suffice to get a taste of it: memes (from memory or mimesis) stand in analogy to genes. They are units of culture that can change in frequency. And that is then roughly all that is common among so-called memeticists. But often one finds additional claims, in particular the claim that memes are like genes both ontologically (it is claimed to be a particular thing that produces copies of itself) and explanatorily (it is the properties of the meme that explains why it spreads).

3:AM: You’re critical of this approach to explaining the mind and culture aren’t you? Why don’t you think memes do what they’re supposed to do?

MK: Memetics reduces again either to something trivial – no big news that there are cultural items that spread, right – or to something false: that they make copies of themselves and that they spread because of their properties. The last point, an explanatory analogy, is often taken to also imply that memes spread independently of the beliefs and interests of human beings, which is also wrong, I claim. It is true: humans learn ideas from others, but again: we do not need memetics to tell us about this. In addition, ideas – call them memes if you like – spread according to a Darwinian pattern. Again: no big news and sometimes false. But that pattern of variational change does not necessarily involve copying (or replication as the more technical term goes), if we try to be precise about what replication means.

Finally, it is still humans that select ideas. They do so, true, in part because of the properties of the ideas, but never independent of the properties of the mind that selects them. Talk about ‘egoistic memes’, I argued, contains a serious mistake, namely falsely applying the Darwinian ontology to the case of culture. The organism in biological evolution is not simultaneously the environment that does the selecting of traits (and ultimately of genes of that organism). Yet, the mind in cultural evolution is simultaneously the entity that does the selection (environment) and the entity in which the selected items (‘memes’) express themselves (organism). In the case of mind, the environment is the organism. A case of auto-selection, as I call it.

3:AM: Memes are sometimes presented as part of the Darwinian assault on our self image. You don’t see Darwinian theory as doing that though do you – you see it more ‘like a disorientated sheep in wolf’s clothing’. What do you mean? Are you saying that reducing notions of ourselves to Darwinian processes is not a threat, or that we can’t be reductionists?

MK: With the ‘disoriented sheep in wolf’s clothing’-claim I tried to say that the aura of danger, which is created with the language of ‘egoistic memes’, is just pretending. It is pretending to be a wolf since pretending to be a wolf – communicating that there is a dangerous new theory – certainly sells much better. It is sad that academia right now more and more takes over the logic of the ‘breaking news’ market. But the truth does often not align with the current mass media’s logic of shock, awe and paranoia.

3:AM: This leads on to the question about human nature. You’ve looked at traditional notions of human essence and what to do with the idea since Darwin came along. So firstly, what options are there given the current Darwinian anti-essentialism?

MK: There is a broad consensus among philosophers of science that Darwinism is incompatible with essentialism. For a Darwinian, essences in the traditional sense of the analytic philosopher do not exist. One option is then: so much the worse for Darwin. If philosophy (e.g. because of a belief in essence) conflicts with science, then science looses. Metaphysically minded people in mainstream analytic philosophy tend to go for that option. In effect, one then says: there are essences despite Darwinian anti-essentialism. Another option is to bite the bullet and conclude that there are indeed no essences and with the essences the concept of human nature has got to go too, since nature mainly meant essence. Finally, you can take a revisionary stance: revise the concept of essence and nature so that they become compatible with Darwinism.

3:AM: Are you arguing that there is a plurality of concepts of human nature that are legitimate so long as we’re clear what role they have in which particular and clearly defined context?

MK: Yes. If you take a revisionary stance, as I did with my collaborators Neil Roughley and Georg Toepfer in our article on human nature beyond traditional essences, you see that indeed there is something in the world that can be called a human nature. The world is such that there are descriptive generalizations about the human life form, describing simply how we are. We call this the ‘descriptive nature’. There are also grounds for explanatory generalizations, referring to a set of developmental resources (not all innate!) that explain why we are the way we are. We call this set of explanatory generalizations (and what it refers to) ‘explanatory nature’. But, neither of these sets of generalizations gives you criteria to decide who belongs to humankind. An individual being might be a human even without having the properties that belong to the descriptive nature (e.g. a disabled person sitting in a wheel chair missing the upright gait that biologist regard as uniquely human and thus part of human nature).

There can also be an individual being that is the perfect example for a human being (regarding the properties that belong to the descriptive nature) but is not born to humans. Imagine that it came into existence in a different way, e.g. via a Frankenstein-like synthesis out of chemical material. Conceptually, this is possible. Many people, we assumed, would regard such an individual being as a person, but not as a human. So, where does that leave us? First, there is a concept of being human that has to do with genealogy: where you come from matters for who you are. Second, the descriptive nature does not give you conditions that allow you to classify individuals. But that is what an essence (as traditionally understood) did: it gave you generalizations about a kind and at the same time a definition, i.e., necessary and sufficient conditions for membership in the kind. An essence simultaneously fulfilled a classificatory, explanatory and descriptive role. In the post-essentialist frame, these roles are still fulfilled by something in the world, but it is in each case different things in the world that fulfill these roles: there is a genealogical nexus between individuals (i.e. a genealogical relatedness between individuals) that allows to classify humans into one big ‘family’ (a classificatory nature); there is a statistically describable set of traits (physiological, behavioral, mental) that describes how humans are (descriptive nature); and there are generalizations that explain why humans are the way they are (explanatory nature).

That’s all. Essence is dead, but human nature concepts (in the plural!) survive. From this, we concluded, that as long as you make explicit which of the three post-essentialist natures you refer to, it makes sense to talk about human nature.

3:AM: Not only have you looked at change as understood in culture but you’ve also looked at how different sciences look at change differently. A striking field is medicine where you say there is an interactionist consensus that conceptually leaves room for a ‘genetic determinism’. Can you outline what you mean by this and how it connects with the innate/acquired distinction?

MK: The so-called ‘interactionist consensus’ says that the ontogenetic development of traits involve an interaction of ‘nature’ (roughly, biologically inherited, i.e. innate factors) and nurture (roughly, everything else that can influence development, i.e. that is acquired). This claim, as I try to show in my current research, embraces five separate claims that need to be distinguished clearly. But irrespective of the distinctions that one could make, it holds that it makes sense to distinguish between nature and nurture, i.e. to treat them as separate kinds of causal factors. This claim can also be summarized as saying that in order for causal factors to interact they need to be separate ‚difference makers’. If we treat them as separate, genetic determinism in a strict sense is on the one hand conceptually possible, but on the other hand unlikely to be found in nature and in any case hard to prove. In its strict form, genetic determinism states that whatever we do with nurture (e.g. whatever change we introduce with respect to nutrition or any other factor), this does not make any difference (or change) with respect to the trait at issue (e.g. body height).

Such a situation is quite rare, as the measurements in the fields and labs of biological sciences show, e.g. with respect to plants. In addition, and most importantly for the implications of all this for social issues: even if it were the case that the change in nurture is not making a difference in the measured situations, one cannot extrapolate to new situations. In other words, if our social programs (e.g., to educate those ‚left’ behind) were not successful so far, we cannot infer that nothing would be successful. We cannot infer genetic determinism in the strict form. So, again: either you say something trivial or something that is wrong (since it cannot be given a truth-value). That genes determine behavior (there are genetic factors involved in the development of this or that) is genetic determinism loose sense and quite trivial. To claim that other factors are not having a potential to impact the trait is genetic determinism strict sense and cannot be proven to be true.

[R.G. Collingwood]

3:AM: What are the stakes in your discussion of causation and Collingwood’s distinction between ‘ability to control’ and ‘willingness to control.’ Can you first sketch out for us how you understand the distinction.

MK: Collingwood claimed that in practical contexts (where ‘setting things right’ is at issue, such as in applied science, medicine, law), we choose among the ontologically given causes (of which there are many) and select some of these for inclusion in a causal explanation. Which one(s) we select depends on our ability to control. Here is an example that he uses. Imagine your car stops at a hill. You wonder why. Now, even though you could also consider the hill as a cause (since it is a cause in the sense that the hill makes a difference to whether or not the car moves), you only consider the innards of the car (e.g. a loose cable) as a cause of the stopping. This is so, according to Collingwood, simply because a typical car driver cannot intervene, i.e. have control over, the hill.

The problem with his account is that there are cases in practical contexts where certain causes are foregrounded (i.e., included) in an explanation despite the absence of an ability to control, and there are cases where there is control but the causes are backgrounded (i.e. excluded from the explanation). I use genes at the beginning of the 20th century as example for the first kind of troubling case (inclusion but no control) and environmental factors as an example for the second kind of troubling case (excluded, despite controllability). So, the ability to control fails to cover these cases of selectivity in causal explanation.

But a slight revision of Collingwood’s principle is sufficient to cover these cases: it is the willingness to control not actual controllability that explains the cases of causal selection at issue. I then use the history of shifting explanations of cancer incidences (as due to factors intrinsic to individuals or as due to environmental pollution) in the 20th century as further evidence for that claim. The resulting picture is: depending on who you are, you are prepared to intervene in some but not other causal factors, and therefore you are selecting a different cause for inclusion in an explanation. The pollution producing industry, for instance, had no interest in intervening in that causal factor when increase in cancer susceptibilities became an issue in public health debates in the second half of the 20th century. Thus, a public health system that aligns with the interests of that industry has no interest in intervening in pollution production and will thus rather explain the increase in cancer incidences with reference to improper, unhealthy behavior of individual beings.

3:AM: How does this show that norms make causes?

MK: What you are interested in to change (or keep as it is) depends on your individual values, but it can also depend on the norms in a society. These norms (and the values connected to them) influence, first of all, which causes are included in an explanation. For instance, if consuming is of high value in a society, then the norm is to follow that logic of consumerism, which will make it unlikely that the industry producing consumer goods will be intervened on (since it is constitutive for the norms of the society). This then, as described, impacts which causes are included in an explanation. This is how norms make causes visible. But this selectivity in explanation-giving can also influence how humans behave. If it does, that can create the causes that are available for inclusion in an explanation.

This looping effect (a term I take from Ian Hacking and adapt it to the case of explanation) from explanations to behavior can go in two directions: negative feedback (the cause becomes less frequent or less strong) or positive feedback (the cause is made more frequent), depending, for instance, on whether the phenomenon we want to explain is agreed on to be something to be prevented or to be furthered. If we (as a society) included the environmental pollution in our explanations of cancer incidences, then that cause can as a result – over time – become less frequent, or at least less strong. If we include education in the explanation of literacy, because we normatively agree that we are willing to intervene on that (e.g. spend more money on education, so that no child is left behind), then that can make that cause more frequent or at least more strong. The message is: we change the world (which causes are there) via our selective explanations. That is how norms make causes real (rather than only visible).

3:AM: Causes are often complex affairs. In order to grasp them, we might decompose so we get to the parts of the whole so we see what’s going on clearer, or we might go up towards a greater level abstraction. So why is it important to understand that the phenomena we’re trying to understand are moving targets and what is going on when we go all abstract, where someone might argue that you start to lose contact with the phenomena?

MK: The problem is: what are phenomena? The way I see it includes the following: the world (if I am allowed to assume that there is one, whether or not we can directly get access to it) can be enriched with phenomena, depending on disciplinary interest. The example I use is body height, a trait of individual organisms. If you take this as the phenomenon to be explained than you have a lot of causal factors influencing it. Given the account of causal selection that I described above, you select the cause you are willing to intervene on. Genetic factors, for instance, are causes geneticists are – by disciplinary affiliation – willing to control, i.e. to intervene on experimentally, in order to give their scientific explanations. As a complementary strategy, you can also adapt the phenomenon so that you get more exclusivity for your preferred explanation.

Assume that there are convincing scientific data that show that women and men differ statistically with respect to their body height. That statistical difference with respect to body height (of all men taken together versus all women taken together) is abstracted from the actual body heights in individual organisms. The trick is now: that difference – an abstract phenomenon – might well stay the same irrespective of what you have done so far with the environment of human organisms (e.g. change the nutrition). So, for the tested environmental changes, the environment is not making a difference. Only genetic factors are making a difference since women and men have statistically viewed different body heights. Then you add a name to that statistical difference, e.g. by calling it the ‘gendered height effect’, and you have reconstituted your phenomenon so that you have more exclusivity for your preferred causal factor. The gene is then the only difference maker for that phenomenon (given measured environments), even though the environment is still a difference maker for body height, the phenomenon you started with.

The analytic point is: phenomena are moving targets and can lead to a lot of talking past each other if it is not made exactly clear what the phenomenon is that is explained: body height or a difference with respect to body height? The more general and critical goal of my analysis in my paper on reconstituting phenomena was to show that one can fiddle with an explanation not only via causal selection (or by fiddling with data, a frequently addressed phenomenon in philosophy of science), but also by fiddling with the phenomena involved. Reconstituting the phenomena can be a sleight of hand.

3:AM: Your work is interdisciplinary. You’ve written about the problems of interdisciplinary philosophy of science where a practitioner is expected to be up on the philosophical literature and up on the scientific research as well. Is this situation recognised by the universities and research institutes or is the culture in those places still biased towards a mono-disciplinary approach despite everything? If it is a problem, is there a solution?

MK: It depends on the structures in the academic institutions. Some are better, some are worse on this. The problem is that interdisciplinarity is very much welcomed in results but not furthered enough in its emergence. It is hardest for young researchers since they are the most vulnerable ones. They are asked by their supervisors to be as interdisciplinary as possible, but once these PhD-students apply for jobs, they have it much harder to get one. The jobs (especially the permanent ones) are still often filled in a rather conservative, risk-aversive, disciplinary manner. And this is expectable if you look at the structure of academic institutions, especially universities. If you have to fill a job in a department that educates students along disciplinary lines, you need to make sure that the program is running ‘as normal’. Even if you would like to, you often cannot afford the risk to hire somebody with an interdisciplinary profile, be this because the department is small, or because all the metrics involved in financial decisions are biased towards disciplinary profiles (e.g. via citation circles). Another issue, it occurs later in the career, is that when you apply for external funding there are sometimes no interdisciplinary committees in the funding body that can evaluate the proposals as fair as a disciplinary committee can for the disciplinary proposals. If you talk to colleagues, most agree that if you do a more conservative topic that fits the well-beaten disciplinary tracks you have a higher probability of getting positively reviewed. The universities, funding bodies and also the referees try hard to counterbalance this, but they struggle with something that is so entrenched that they often still fail, despite best intentions. Nonetheless, my impression is that the situation is much better than 20 years ago. This probably holds for all issues about diversity in philosophy.

3:AM: And finally for the readers here at 3:AM, are there five books that we should be reading to get further into your philosophical world?

MK: Sure:


• Ernst Cassirer, 1944, An essay on man.

• Stephen Downes and Edouard Machery (eds.), 2013, Arguing about human nature.

• Evelyn Fox Keller, 2010, The mirage of space between nature and nurture.

• Peter Godfrey-Smith, 2009, Darwinian populations.

• Sandra Mitchell, 2009, Unsimple truths.

ABOUT THE INTERVIEWER
Richard Marshall is still biding his time.

Buy his book here to keep him biding!

First published in 3:AM Magazine: Saturday, January 14th, 2017.