The Cliology of –
Cognitive Gadgets: The Cultural Evolution of Thinking
I first heard Cecilia Hayes at her keynote lecture at the inaugural meeting of the Cultural Evolution Society in Jena, Germany in 2017. She is Seinour Research Fellow in Theoretical Life Sciences at All Souls College Oxford: a professor of psychology, and talks with a cut glass received English accent that would be ideally suited to a BBC Radio 4 newscaster. I was captivated by her ability to talk for two hours without going “erm…” or “er…”; her writing style is equally eloquent though sometimes overly academic. Anyway, to the literature content.
Overview
As to what makes humans unique among the species (or at least think we are) has long puzzled mankind. Hayes goes beyond gene-centricity to argue that we possess “cognitive gadgets”. These are acquired thinking tools and products from cultural evolution.
This view is clearly related to memetics and cliology, and the diagram on page 35 pitches cognitive gadgets theory within the selectionist conception of cultural evolution, either alongside, or possibly containing, memetics (but not within memetics). This proximity makes the idea of cognitive gadgets an attractive source of inspiration for cliology, and thereby merits a more thorough treatment. The term “gadget” itself has engineering connotations. I hope Cecilia will forgive me if I occasionally abbreviate this to “cogs”, a familiar metaphor even if it does risk reducing it to some Victorian clockwork mechanistic worldview.
Of course, the nature vs nurture debate is as old as the hills and since the unravelling of the structure of DNA, the gene-centric view has become equally tardy. However, Hayes throws a refreshing new light onto the subject by arguing that, although we do have cognitive instincts determined by our neuro-anatomy, and this does present a platform upon which we run our particular thoughts, there is an intermediate layer of culturally acquired know-how in between – cognitive gadgets. If we were talking about computer hardware and software, then cognitive gadgets would be the equivalent of install routines.
In this view, cultural evolution has two levels: the substance of culture (or grist in Hayes’ metaphor) is about routines, skills, technologies and so on. On the other hand, there is the processing of this grist using mills – the meta-skills of acquiring skills. However, according to Hayes, these mills are not anatomically predetermined by the genes, but are themselves evolved elements of culture.
It may seem fairly obvious that individual languages have evolved over time, and that a child learns to talk, read and write through exposure with those who already have learned the language. It is tempting to think that, while the language itself may consist of arbitrary cultural conventions, the child’s acquisition of language is instinctive and is ruled by the genes. Not quite so, claims Hayes. Genes do play a part, but they are not the whole story. Rather than the child’s pure instinct to learn, culture has evolved mechanisms by which to encourage the child to learn, and these cultural mechanisms are passed down (with modification) from generation to generation. Cognitive gadgets then, are evolved cultural components that facilitate what, at first, may appear to be instinctive. They confer fitness not only for individuals and society, but also for the cultural ideas themselves. Hayes does mention memes but shys away from going as far as I would to consider higher containment level ideas as epistememes.
Dissection
ch 1 a question andmany answers
It is an age old question “what makes us peculiar?” Human distinctiveness has four complementary foci, bodies, brain behaviour and mind. Hayes sets out to position the cognitive gadgets theory, drawing on a geographic metaphor, in differentiating it from other theories that have been set out. Chronological approaches provide narrative theories, the key events in evolution, while force theories focus on the processes involved. These historical and forces approaches are two dimensions that provide her coordinate system for her landscape, and upon which she locates the contending schools in order to make comparisons. It also makes for four quarters. The ideal approach, she posits, would be that quarter that combines both history and evolutionary forces.
Cognitive gadgets is a force theory: it considers the forces that shape evolution. It applies cultural evolutionary theory to the mechanisms, the forces that shape thought (p13). The distinction is made through metaphor; that cognitive gadgets is not concerned with the “grist” of the mind – what we do and make – but with its “mills”, the way the mind works. (p14) In contrast to evolutionary psychology, which assumes that the mechanisms of though have been fashioned by the genes as an instinct to learn socially, cognitive gadgets argues that social learning is, in itself, a product of cultural evolution and is not all in the genes.
There is an overlap between Cultural Evolutionary Theory and Evolutionary Psychology. Whereas the “all in the genes” position of Evolutionary Psychology provides a wholly instinctive mechanism behind culture, Cultural Evolutionary Theory has it that “traits” can be passed down to cultural descendents who may not necessarily be genetic descendents. But cultural evolutionary theory says little about the actual mechanisms of inheritance. The intersection of these areas is one that agrees about non-genetic trait inheritance, but does not consider that the mechanisms are entirely instinctive nor genetically predetermined. Genes do play some part, but it is not “all in the genes”. Rather, the mechanisms of inheritance are themselves, products of cultural evolution. From this composite view we get, what Hayes calls “cultural evolutionary psychology”.
This hybrid term is apt in denoting the interdisciplinary ground that marries the two fields, yielding a non-gene centric approach to the mechanisms of social learning. It also offers something of a container field, a potential niche academic haven for those who struggle with reducing the processes of cultural evolution immediately to the genetic level. It is a candidate home (or at least a B&B) for cliology, even perhaps memetics.
Notwithstanding, Hayes’ landscape and positioning therein of her cognitive gadgets theory (as cultural evolutionary psychology), allows us to locate and thereby make comparisons with its neighbors of evolutionary psychology; folk psychology; cognitive science; cultural evolution with the maxim that “culture changes the way we think”.
——-
ch 2 nature nurture culture (review on p51)
A core question is: for any particular characteristic, to what extent *and* in what ways do nature and nurture contribute to its development. Nature here, can be seen as genetic inheritance of characteristics. Nurture, on the other hand is not genetically inherited, but rather these nurtured characteristics arise through direct interaction with the environment. Cultural characteristics come from a cross between nurthred characteristics as acquired from others in the environment, whereby those nurtured characteristics are culturally inherited.
Neither nature nor nurture alone contribute to development, rather, characteristics arise through the interplay between them. We require a common currency by which to make comparisons, measure the influence of genetic or environmental factors, and avoid partisan theories of nature vs nurture. The commonality between genes and culture is that they are both essentially modes of transporting information. The teleo-semantic view, which uses the concept of Shannon information (correlational), can be applied to putting nature and nurture on a comparable par – that of information – such that they can be understood in the same way.
Any trait, cultural or biological, that correlates to an event and was caused by that event contains information about that event. Hence certain traits hold information about events, and it is this information holding capacity that allows us to apply the same measure to nature nurture and culture. Biological and cultural evolution, when seen as information, can also be made comparable.
P30 cultural ev theories:
Cognitive Gadgets is a selectionist theory of cultural evolution, but to put that into perspective, Hayes surveys the range of cultural evolution theories along with their features and strengths. The “historical” perspective is loose as it simply provides a narrative of events. The populationist view considers cultural characteristics that are acquired through kinds of social-learning and can be thought of as evolutionary for three reasons: social learning built on a set of genetically inherited cognitive instinct for learning from others; it takes a “dual inheritance” stance of coevolution between socially acquired characteristics and genetically inherited characteristics; and it is methodologically evolutionary as it applies the mathematics of population genetics to socially acquired characteristics. The selectionist view goes on to argue that cultural evolution is Darwinian but that inheritance mechanisms are not genetic.
One step further, and one that is clustered by Hayes into the selectionist camp, is that of memetics.
… most scientists and philosophers doubt that memetics is sufficiently coherent, or grounded in empirical evidence, to provide a basis for a research program. (p35)
Memetics, being one of the key metaphors of cliology, is central to this review. It is worth considering why Hays does not do more with memetics, and further, what cognitive gadgets would offer had she have done so.
What she does persue p36: Cambellian selectionist culture and genes share the same fundamental houristic of blind-variation-and-selective-retention
P34 campbell selectionist darwinian blind-variation-selective retention.
inheritance mechanisms are cultural rather than genetic.
memes dawkins dennett
Arguably, this is the reason why, after thirty years of conceptual development, memetics still has not been converted into an empirical research program, and its hypotheses still rarely inspire observation and experiment.
So, coherence and grounding in empirical evidence being the basis for a research programme. What is she missing from saying this?
The point of memetics is that “memes are in it for themselves” but might gang up with other replicators for mutual self-interest.
However, it suggests that , rather than being on a short “genetic leash”, cultural evolution is highly autonomous with respect to genetic evolution.
Hard to unitize psychological grist – ideas and behaviour
P37 The mind, partitioned by the “mills” and “grist” metaphor is used to say that evolutionary theory explains not only what the mind works on “grist”, but how the mind works “mills” from an evolutionary perspective. In other words, not only the products of culture “grist” can be explained in evolutionary terms, but also the processes “mills” can be explained in evolutionary terms.
To apply cultural evolution to the mills, is similar to the problems encountered in memetics and we need to ask: What entities are evolving? What are the routes to inheritance? whether vertical, oblique, horizontal or a combination thereof, and what kind of social learning that provide an inheritance system?
P38 Hayes suggests that the problem of what entities are evolving is a possible reason why memetics didn’t take off. The quantification and unitization of psychological grist is difficult (ideas and behaviour are identified by folk psychology rather than scientific taxonomies and markers), which has impeded an empirical research program into the content of culture.
On the other hand, mills can be unitized by cognitive science, thereby identifying distinct cognitive mechanisms that are subject to darwinism. These gadgets have recognisable traits such as functional signatures, because they are algorithms.
This turn of direction demands comment. To elect not to employ meme theory seems mostly down to the state of play of memetics as a science. The unitization of psychological grist (ie memes) has not progressed beyond folk psychology. It is hard, but not impossible, it just means that the tools remain to become more widely available and the results more widely disseminating. Cognitive mechanisms were once at the folk psychology stage, but cognitive science has managed to overcome an impasse; one that still evades memetics. Hayes’ field requires a platform on which to build, rather than having to build that platform, which is understandably outside her role. Were is that the tools for unitization of grist available, and I believe that Cultural Linneanism may lead to such, then that would release memetics from its quagmire and allow empiricism of cultural evolutionary psychology to work on both the mills and the grist.
Ch3 starter kit
Humans acquire much knowledge and skill over a lifetime. But we do not start with absolutely nothing. We have an innate predisposition to make such acquisitions: a genetic starter-kit, so to speak. Evolutionary psychology says that the genetics that bestow our psychological attributes would account for the big special differences between human and non-human. Our biological uniqueness is what allows for the acquisition of culture.
The cognitive gadgets view, on the other hand, says that these psychological attributed, as endowed by our genes, are not the big special in all this, but rather, just the small ordinary. That is, our differences to other primates are not that great, but it is those minor differences, the small ordinary in psychological attributes that amplify into the big special of cognitive mechanisms. Small ordinary changes can have major downstream effects. These effects are the distinguishing features of humanity, but they are arrived at through non-genetic means. The big special, the cognitive mechanisms, then are an intermediary between genetics and the details of culture.
The small ordinary genetic predispositions that allow for the big special non-genetic cognitive mechanisms are:
Temperament: humans are genetically wired for greater social tolerance. Evolution has tweaked this tendency making humans more inclined than our primate ancestors to enjoy response-contingent stimulation (motivation) and therefore to approach and learn from interaction with other agents. These are necessary for effective tuition.
Attentional biases are towards other humans: we have an inborn genetic inclination to facial and voice recognition.
Cognition: our larger brains are capable of capturing a torrent of information but use it to build domain-specific cognitive big-special mechanisms designed by cultural evolution.
Dual process theories associative learning (system 1) is more advanced; but our more advanced prefrontal cortex gives executive function (system 2) — trainable therefore allowing learning and cultural inheritance in executive function. Allows “upload of mills”.
Ch4 Cultural learning
Cultural evolutionary psychology is both a hypothesis and a framework. It proposes that cultural inheritance played a dominant role in shaping the cognitive mechanisms that make humans distinctive. It is argued that the cognitive gadgets view rather than that of cognitive instincts, is supported by the cognitive science data.
Hayes elected to examine four types of “cultural learning”: selective social learning, imitation, mind-reading, and language.
The prevailing consensus among academics is that cultural learning mechanisms are genetically inherited. A critical test is posed:
‘Therefore, in the interest of scientific progress, cultural evolutionary psychology warrants pursuit as a descendant of, and alternative to, evolutionary psychology and cultural evolutionary theory only if it can challenge this consensus.’ (pp 78-79)
A long passage quoting Henrich typifies how cultural learning is usually thought of, but its shortcomings are called out. In particular is the issue that it ‘does not offer a clear and coherent way of distinguishing cultural learning from other kinds of social learning’; claims of complexity are obviated as it does not give any dimension by which it is especially complex; and that ability to make inferences (about preferences, goals, beliefs or strategies) could be taken in a strong sense of rule governed learning (too high a bar) or a weak sense which fails to discern from any other learning.
An alternative which distinguishes a hierarchy of learning styles, individual, social, or cultural learning concerns not as what is learned but how it is learned. Social learning is a specialised form of individual learning (the individual is doing the learning) involving others through imitation and so on. Cultural learning, further, is a specialised form of social learning, and therefore an even more specialised form of individual learning, that is cumulative and begets cultural inheritance. These are not partitioned, but rather more derived cases of a more general style of learning. This allows for the same cognitive mechanisms across all learning.
Ch 5-8
The next four chapters deep-dive into distinctly human cognitive mechanisms and are presented through case studies (which are at a level of detail beyond the scope of this review): Selective social learning, Imitation, mind reading, language. The argument put forward is that these are cognitive gadgets rather than cognitive instincts. And no longer in the genes alone
Ch 9
The final chapter examines cultural evolutionary psychology. This blends cultural evolution with evolutionary psychology to address the question of what makes us distinctly human. To avoid any uncertainty over selectionism, her view is made clear in claiming that distinctly human cognitive mechanisms are products of cultural evolution and shaped by group selection. All selection processes require a mechanism of inheritance.
However, inheritance of cognitive mills can also involve group-level or social processes – things that go on between people, rather than inside individuals’ heads – such as conversation, storytelling, turn-taking, collective reminiscing, teaching, demonstrating and engaging in synchronous drills. p203
Memetics may be stigmatised, but she is almost saying that these mechanisms of inheritance are memetic: ideas are passed between people.
A fairly common way of passing on information is to teach what should be done, but omitting how to do it or how it should be done. That is, to provide many examples and feedback, but the learner has to figure it out the actual skill for theirself. Face (and wine recognition) recognition is a key skill whereby the nack has to be derived by the individual (p 220)
On the other hand, on top of teaching what to learn, the “how to” can also be taught – this is epistemic engineering, of which literacy training is an example.
… this does not occur by chance or as a consequence of blind selection process. Literacy training is intended to make children literate; it has been designed by educators to have exactly that effect. P205
Epistemology, being about how we know what we know, makes epistemic engineering an order above simply declaring what is important to know or the importance of being able to do something. Teaching the importance of correct spelling may be valuable, but just telling the learner what words to spell leaves them left to their own devices and they might then learn how to spell phonetically – which happens to be a lousy way of learning the spelling of words and can lead to literacy impediments and avoidance of education (as I experienced first hand). A good speller often visualises the word, so good literacy involves teaching visualisation techniques. For many areas though, the teaching of effective systems is presently missing.
… in the future, the cultural inheritance of other cognitive mechanisms could be enhanced by formal education. It may be possible … to improve cognitive skills such as selective social learning, imitation, and mind reading in the whole society or particular group. In these cases, as with literacy and numeracy, cultural selection could be augmented by intelligent design. pp205-206
To add to this notion, there is a possibility of a higher-order learning which would not just teach how to learn that specific thing alongside what to learn, but provide a more general strategy of how to acquire learning skills (including, and recursively, the acquisition of learning communication and teaching skills). Such meta-learning would allow much more strategies for learning to be derived and communicated through culture.
The mantra of evolutionary psychology is that of “our modern skulls house a stone age mind”. Notwithstanding that we might simply be living in the latter part of the stone age, the inference is that our contemporary technological culture is advancing at a pace that our genetically evolved brains are struggling to keep up with. It might often feel that way, but the collective intelligence hypothesis suggests that the essential anatomical capacity to handle complexity (the small ordinary) was already in place long before that complexity emerged. Rather than a change in brain structure, it was demographic changes that allowed for the big special. Networks of social structures reached some kind of critical mass, a tipping point, whereby the interconnectivity enabled an explosion in cultural evolution. p212
Cultural evolution – in the strong, selectionist sense – has produced the hidden causes, not just surface manifestations, of human distinctiveness.p216
Hence, unique human nature is not owed to any radical genetic evolution of the human brain, but rather modernity stems from the coming together of our stone-age minds. If evolving culture invokes human nature, then human nature is also highly mutable. A “phylogenetic tree” of speciations and extinctions of many human natures might be imagined whereby the success of any species of human nature is subject to selection pressure. This could be bad news. While some catastrophe might not extinguish us humans as organisms, it might be an extinction level event sufficient to cull our current nature, plummeting us straight back into the stone-age. (217-218). On the other hand, this agility means that human beings can rapidly adapt our human nature to meet the demands of technological progress.
rather than taxing an outdated mind, new technologies – social media, robotics, virtual reality – merely provide the stimulus for the cultural evolution of the human mind. p218
Four case studies were presented covering selected cognitive gadgets, but a fifth type, that of moral or normative thinking, is proposed. A cognitive gadget that drives what we “ought to think” is an interesting point for memetics because it promotes the fidelity of cultural inheritance. A cultural norm or taboo is one that propagates itself and suppresses competing thoughts. It also acts as a filter, an epistememe, that allows memes that support it to be accepted, or rejects those that undermine it. The content of some specific norm then is the nature of the culture; as a cognitive gadget, normative thinking is what allows such norms to become widespread as culture. So far this gadget has not been tested empirically, but it does suggest that other cognitive gadgets may exist. Furthermore, there may exist such other things as emotional gadgets 220? The call then is to ‘test each distinctly human cognitive mechanism for “gadgetry”.’ First, however, it would be useful to draw up a checklist of features that indicate that a mechanism is a candidate for being a cognitive gadget.
In conclusion, “It’s in our genes” cannot be used as a default explanation anymore, the distinctly human cognitive processes are products of cultural group selection . they are not cognitive instincts, but cognitive gadgets.223
Clioanalysis
Cicellia Hayes asks what is it to be distinctly human? Cliology asks, what can we learn from Cicellia Hayes in order to toolup for a better, less maladaptive, culture. At first, there seems to be no real link between these questions as, that which separates us from beasties is mostly academic. It doesn’t particularly progress towards engineering solutions. However, the link comes more into focus in positing the relationship between cognitive gadgets and memetics.
Understandably, Hayes, being a respected academic, is reticent to build upon memetics as it is still stuck in a methodological quandary. As the task of cliology has a futurealogical temperament, of forecasting and fabricating culture, then the task of speculation of how cognitive gadgets might be applied for cultural benefit, given that some derivative of memetics does have empirical potential, falls to reviews such as this one. Of course this is all guesswork, as with any forecasting, but it involves scrolling through the theory, searching for possible tools that might coherently lead somewhere.
A big chunk of the book is devoted to furnishing the evidence for cognitive gadgets (Ch 5-8) while the shorter final chapter is more exploratory. This is safe, academically, but the emphasis of this review, in projecting these lessons into application, carries an inverse emphasis – the final chapter is where the meat lies.
In summary, cognitive gadgets are what makes us human: they are the mills that gives us the processed grist – the mechanisms that generate the content of culture. A subsequent issue arises: as human-culture interaction is unavoidable, then how can we apply this to making us better at being human?
Ethics, moral content, and whatever “better” means aside, what tools does this give us? Firstly, there is some risk of confusion about tooling levels. We are not talking particularly about flint axes or microchip technology here; they are of course tools at a user level, but they are the grist as processed by the mills. The level the tooling of interest are the mills – these mills are meta-tools. Cognitive gadgets, it is argued, give “better” culture and therefore “better” human nature: a more modern modernity (given that the word modern is derived from adaptive). So, although not necessarily, better grist processing allows for better user level tools and techniques.
To address cultural relativism in this we might ask “what have the Romans ever done for us?” Modernity has its blessings, however mixed, and even given the cold comfort that we might eventually recover from being plunged into a post apocalyptic stone age, we might gather that cultural progressivism might be desirable, as we have hardly attained the peak of cultural enlightenment. This form of progress might give us better technical toys, but the exciting prospect of novel cognitive gadgets, built of intelligent design is the aspiration of “better living through cliology”.
What might this level of epistemic engineering entail and what might the application be? Would it be possible, through intelligent design, rather than blind variation and selective retention, to construct a mill to specification; from scratch? Well, we have to revisit the distinction between science and engineering. The science of psychology attempts to understand the mind through study which is clearly Cicelia Hayes’ area. Pragmatists want to apply that understanding, and often in their frustration over what they see as pedestrian progress in research, leapfrog some of the more pedantic points. This can lead academics to conclude that the models pragmatists are attempting to build are pseudo-scientific. To be sure, they are not intended to be scientific, but rather working engineering models that bestow results. Neuro-Linguistic Programming (NLP) has been lambasted for a variety of reasons, but mostly as a pseudo-science of psychology. A lot of this might be down to the misapprehensions of received wisdom, or at least the way the courses were marketed. Were NLP to be seen solely at the level of psychotherapy, as some commentators do, then the consensus over concerns of empirical credibility would be valid. However, pragmatists strive for results, rather than the detached study of results as a means of evaluating theory, and personal anecdotes are the measure.
This is not to be apologetic about NLP, but rather to draw attention to its relationship with epistemic engineering. NLP appears to be pseudoscience from the lense of academic psychology, which focuses on the grist of therapeutic intervention. NLP though can be seen as an attempt at building a framework for epistemic engineering; it is a mill, a cognitive gadget. NLP, itself, has little to do with psychotherapy, rather, it is a coding convention, a programming language, whereby the behaviour of one person might be transferred to another.
As with face (and wine) recognition, we know what things are important to learn, but are left to our own devices in figuring how to learn those things. Literacy training, on the other hand, has been deliberately codified in a teachable way – it is a product of epistemic engineering. We learn much through imitation, but are not taught how to imitate. NLP essentially has codified methods of imitation such those literate in the language of NLP can use can intentionally acquire skill sets more effectively. NLP might therefore be seen as the intelligent design of the cognitive gadget of imitation. NLP also provides a language for encoding and transferring such behavioural programming, and as such, is a programming language. Programming languages for human or machine behaviour sit within the domain of engineering rather than scientific study: they are about practicality. If NLP was developed today, it might well be known as Human Behavior Mark-up Language (HBML), much like HTML (Hyper-Text Markup Language) which is used to encode web pages. Psychotherapy is an attractive skill to be able to transfer from one person to another, but much like a web page on the subject of psychotherapy, it is not the code it is written in – neither HTML nor NLP are psychotherapy.
NLP, whatever one thinks of it, was chosen to illustrate that other examples of intelligently designed cognitive gadgets are likely. It also points out the distortions of confusing mills with grist: the user level tools from the meta level tools. From a cliological perspective then, we must ask, like literacy and numeracy training devices, whether cultural inheritance of other cognitive mechanisms could be enhanced by formal education; what other types of cognitive gadget can be built? While the benefits could be debated, the question is can we improve cognitive skills such as selective social learning, imitation, mind reading, and others yet to be identified, across society?
While drilling down into the empirical evidence that supports existing gadgets, it would also be worth forming a structured taxonomy according to traits of current gadgets. Finding the empty spaces, much as with Mandeleev’s table leading to finding new elements, may betray the presence of “eka-gadgets”. Not only would a candidate gap-in-the-map give the psychologists something to look for, but would also give the epistemic engineers structural insight into how gadgets may be encoded and disseminated. Hayes has chosen the metaphor of mills and grist to highlight that cognitive gadgets, while still being parts of culture, mills are of a different order than the grist of most cultural content. This supposes two levels of cultural things, but in doing so raises speculation for more levels, perhaps above or below, or even some kind of continuum in-between.
The point about the nature of humanity (opposed to just Homo sapiens sapiens) as being as much a product of culture as it is the cause of culture, when thought of in cultural evolutionary terms, raises conjectures concerning evolutionary descent and “phylogenetic tree” of human nature (again, as opposed to hominids). There may be a whole Linneanan sort of map of depicting the range of human natures, of which we can fit selective social learning, imitation and so on.
In terms of cliological speculation, this is where we head into ethically dodgy territory. Not only would memetic engineering be applied to the grist, but also the mills – that is, while the genome of our bodies is becoming amenable to intelligent design through genetic modification and vectorology, so too will the fundamental patterns of human nature succumb to engineering principles.
In the cliology of cognitive gadgets, it can be seen that clliology is at the level of cognitive gadget.