In their target article, Osiurak and Reynaud (O&R) emphasize that “technical reasoning” critically contributes – as an individual ability – to cumulative technological culture (CTC). I comment on several of their statements in the context of a novel, coherent theoretical perspective.

I here conceptualize tools to be in a duality with their corresponding mental representations, as kinds of mirror images. Such a view is supported by active externalism (Clark & Chalmers Reference Clark and Chalmers1998; Menary Reference Menary2010) stating that (brain-) external objects are directly coupled to cognitive abilities that otherwise could not happen. For example, mathematics is a cognitive activity that – beyond the limited capacity of mental arithmetics – cannot be performed without tools, that is, external objects.

The term “tools” here is used in its widest sense which includes artifacts. Tools are produced in a defined sequence of observable acts and serve one or several uses. Tools elicit two kinds of affordances: first, those for their use, and second, mental representations on how they were produced; for simple objects, these mental representations may be accurate to detail (“ladder” and “hammer”), but for complex tools these may be diffuse (“computer”). These affordances can be described as mental representations or instructions corresponding to externalized, clearly defined action sequences of observables (see Fig. 1).

Figure 1.

Scheme on the tool–mental representation duality showing the flow between the individual's mental representations and tools in replication cycles. Mutations happen in the internal, private sphere, replication and selection happens in the external, public sphere.

Mental representations can be transmitted between individuals and executed whereby the tool is reproduced. Or they can be varied, including by reshuffling, for example, decomposition and reassembly to novel mental representation sequences. A reshuffled mental representation, corresponding to “technical reasoning,” when executed, will generate a novel, innovative tool. Tools are therefore materialized, or “petrified” memes because the previously modifiable, corresponding mental representations have been fixed and are “archeological” evidence of that prior, observable and discrete sequence.

I propose that such an exclusive and restricted definition of memes in the context of tools and their corresponding mental representations can be a useful, unifying, and fertile concept for bridging macroscale and microscale phenomena, that is, CTC and individual psychology, respectively. Such a definition as a tool–meme duality contrasts to previous fuzzy definitions that attempted to be all-comprising but were ultimately unproductive.

The meme shuttles between a representational stage (coded in the brain by unknown mechanisms) and materialized stage molded in the tool (see Fig. 1). The material stage increases meme fidelity and thereby potentiates its replicating power. Tool evolutionary history can, in principle, be tracked in a pedigree-like manner.

The question is not whether a meme is necessary (Henrich et al. Reference Henrich, Boyd and Richerson2008) but whether it can be defined in a useful way providing an instrument for novel transdisciplinary insights. Universal Darwinism (Dawkins Reference Dawkins and Bendall1983) posits that if an algorithm is active that includes variation, replication, and selection, the emergence of replicators is inevitable.

The definition that O&R use for structure and content corresponds to the classical definition of vehicle and replicator (Dawkins Reference Dawkins1976; Reference Dawkins1982), respectively. The more prior tool-related mental representations (replicators) are present in an individual vehicle the less opaque and the more transparent is a tool perceived.

This is not in contradiction to the possibility that individual “hardware” (e.g., brain wiring) differences may exist. The individual ability to process mental representations will be a sum of both software (relevant memes uploaded) and hardware (individual processing, recombination, or affordance abilities).

Within the dual inheritance theory, scientific evidence and a clear framework have been established to explain mutual influencing between cultural (software/content) and biological evolution (hardware/structure) (Boyd & Richerson Reference Boyd and Richerson1985; Henrich Reference Henrich2016).

Inborn behavioral patterns are more consistent and faster available, whereas acquired behaviors are flexible but prone to eradication. Culture-biology coevolution likely has used both modes and combined their advantages. Useful, acquired behaviors have been fixed to make them less labile (“Baldwin effect”) and inborn behaviors have been replaced by more flexible modes. The speed of such a culture-biological coevolution is accelerated enormously by assortative mating (Creanza et al. Reference Creanza, Kolodny and Feldman2017) and is demonstrated by domestication (Hare et al. Reference Hare, Plyusnina, Ignacio, Schepina, Stepika, Wrangham and Trut2005), or self-domestication (Hare Reference Hare2017; Henrich Reference Henrich2016).

Ultimately, at any given moment, there exists a mosaic of hardware- and software-based processing for the brain with all possible graduations – all simultaneously present but each with different phylogenetic and ontogenetic histories.

The existence of specialized brain areas participating in recalling particular kinds of memes (i.e., for technical reasoning area PF) is conceivable: motor tasks and corresponding mental representations may evolve from simple sequential arrangements to those involving objects, and later tools. During this gradual, phylogenetic evolutionary trajectory, specific brain areas are sequentially involved and occupied.

O&R presume that technical reasoning is able to extract “essences.” “Essences” could be elegantly defined as the explicit perception of overlapping features of a large number of tool-related mental representations.

O&R oppose social versus asocial learning and explicitly name technical reasoning as an example for an asocial skill. This distinction ultimately depends on the definition of social learning (Heyes Reference Heyes2012). Although technical reasoning does not require the presence of other individuals, clearly, all technical mental representations themselves have been “uploaded” at some point in time by social learning (including pedagogy) which is domain-general (Heyes Reference Heyes2012). Once uploaded, of course, mental representation can run off in an asocial mode.

For a comprehensive theory of technical reasoning and innovation, the importance of group size and social communication for acquiring technical skills can also integrated by the tool–mental representation duality: experimental evidence shows that social transmission skills win over individual innovation skills at the group level, as long as there are a few innovators present because technical reasoning is more easily copied than reinvented (Henrich Reference Henrich, Shennan and O'Brian2009; Reference Henrich2016, p. 213; Muthukrishna et al. Reference Muthukrishna, Shulman, Vasilescu and Henrich2014).

In conclusion, technical reasoning within CTC cannot be seen outside an ontogenetic and phylogenetic context involving dual inheritance, and a restricted definition of memes as a tool–mental representation duality may be a fertile ground to elegantly bridge these areas.