Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Terrence Deacon’s achievement in the Symbolic Species (1997) and the inspiration stemming from that book is vast. Among its central and challenging ideas is the idea that the overall change in cognitive abilities from higher animals to man is a semiotic one and its description should include semiotic concepts. Another is the idea that hominid semiotic evolution forms an extreme case of Baldwinian evolution with feedback from the common semiotic capabilities of a hominid group onto the selection pressure on each group individual. The resulting hypothesis claims that the co-evolution of human language, cognition, and brain is far more probable than the idea that language and human semiotic capabilities appeared as an accidental side product of a general growth in human cognitive abilities. And an ensuing idea is that human language capability does not form an isolated brain module but is rather involved in the overall evolution of the human brain, uniting formerly distinct capabilities. Finally, the important idea that central parts of linguistic structure are neither innate nor acquired – but rather constitutes a priori structures of any possible thought and communication – which evolution as well as development are forced to adapt to. This idea offers an important third possibility as compared to the ongoing nature-nurture cul-de-sac. Deacon’s later development of concepts like autogen – the structure of primitive pre-cell proto-metabolism as an important step towards the earliest life and semiotic process – and his idea of alternating maskings and unmaskings of selection pressures in evolution add further conceptual tools to a unified explanation of the sophistication of semiotics within the framework of biological evolution.

Still, a conceptual problem remains in the heart of Deacon’s hypothesis and the whole discussion revolving around it. Famously, he reinterpreted, in The Symbolic Species, Peirce’s semiotics in order to find a conceptual skeleton scaffolding his overall evolutionary hypothesis – and picked upon Peirce’s well-know triadic distinction between the sign reference types of Icons, Indices, and Symbols for that purpose. The icon-index-symbol distinction addresses the difference between signs which refer to their object by means of similarity (icons), by means of actual connection (indices) or by means of the habitual, conventional reference to a general idea (symbols), respectively. This distinction – like most of Peirce’s semiotic distinctions – is not one of three separate classes of signs, like three different biological species. It is not the case that signs fall in three mutually exclusive classes named icons, indices and symbols. They rather form aspects of signs and may co-exist in empirical signs. The relation between these aspects is one of increasing complexity – so that most Indices contain Icons, and so that most Symbols contain both Indices and Icons, or, in any case, involves them in realizing its meaning in the ongoing chain of interpretation and reasoning. Certain scholars, Deacon among them, have attempted the obvious idea of mapping this increase in semiotic complexity onto the increase of biological complexity in evolution – so that purely iconic signs could be taken to precede indexical signs evolutionarily, and, in turn, indexical signs could be taken to precede the appearance of symbols in the course of evolution. The origin of symbols, finally, could then be taken to coincide roughly with the origin of man and the specific, plastic and fertile capabilities of human semiotics – so as to form a virtual semiotic missing link. Hence the very title of Deacon’s “Symbolic Species” book.

Agreeing with the overall program of that book, this chapter addresses what I conceive as an important conceptual problem in the particular idea of mapping the icon-index-symbol scale onto the process of evolution – a problem whose solution, I shall argue, entails an important inversion in the conception of the place of semiotics in evolution. First, the symbol concept as defined in Peircean semiotics does not appear in human cognition and communication only. This might be taken to be a strife of words – maybe Deacon’s reinterpretation of Peirce’s symbol concept “gets it right”, so to speak. But the ubiquity of Peircean symbols in pre-human biology points to a deeper problem in Deacon’s seemingly small but significant reinterpretation of the relation between the three sign types: he takes them to be compositional so that indices are simply a special configuration of icons, and symbols, in turn, are but a special configuration of indices. But you can not compose the direct object reference of an index out of ever so many iconic likenesses. The summing up of predicates never results in an object reference. Existence is not a predicate, as Kant realized. And you can not compose the generality of a symbol out of ever so many singular indexical here-and-now references. The icon-index-symbol triad should rather be conceived of as “physiological” in the sense that icons and indices generally form aspects of symbols. This implies, conversely, that we should not expect to find “pure” icons and indices which will rather be limit categories. This “physiological” conception of Peirce’s triad implies that the most complex parts of Peirce’s basic triads should be expected to be realized even in the simplest biological processes – albeit with a lack of internal differentiation. This becomes clear from the wide applicability of the central sign in Peirce’s semiotics, that of propositions or “dicisigns”, in biology. Finally, this paves the way for conceiving of a more precise conceptualization of the semiotic “missing link” in terms of the ability for explicit control of semiotic processes in human beings. We shall attempt to flesh out this notion of “explicit control” in terms of its products, “hypostatic abstractions”; in terms of its structure, sign governing other signs; and in terms of its process in the shape of turn-taking dialogue, making possible the comparison and tentative integration of Deacon’s hypothesis with Michael Tomasello’s “joint attention” hypothesis.

1 Peirce’s Symbol Concept Revisited

I have discussed the problems in Deacon’s reinterpretation of Peirce’s icon-index-symbol triad elsewhere (Stjernfelt, 2007, chapter 11) so I shall restrict myself to summing up that discussion. Peircean symbols are not restricted to human sign use. Symbols are described as signs which are general as to their object; they possess an esse in futuro, referring to a potential continuity of future objects; they refer to their object by means of a habit, natural or cultural; they comprise icons for their understanding and indices for their object reference; and they have propositions as an important subset. Thus, simple Pavlovian conditioning – dogs acquiring the habit of displaying eating behaviour by the ringing of a bell – will constitute a full-fledged Peircean symbol, not merely an indexical sign as Deacon would have it. It is not, like indexical signs, restricted to the here-and-now of actual connectedness between sign and object. The bell sound is a general type, referring, in turn, to another type, that of eating, a potential multitude of future eating situtations. The here-and-now of the particular bell sound token in a particular instant functions as an index incarnating that general meaning in the actual moment of the ringing. Even a case as simple as E. Coli swimming upstream in a sugar gradient as the result of its registration of molecules displaying a specific active site (Cf. Berg, 1988; Stjernfelt, 2007) must be described as symbolic in Peirce’s sense: it is a habit (acquired phylogenetically, to be sure, in contrast to Pavlovian conditioning acquired ontogenetically), the habit connects a specific, typical aspect of molecular shape with a specific, typical action, that of oriented swimming and consumption. That the molecular configuration of the “active site” is general may be seen from the fact that it is not unique to a specific carbohydrate but covers a long range of different carbohydrates (Cf. Adler, Hazelbauer, & Dahl, 1973). This generality is facilitated by the chemoreceptors of the cell being geared to detect sugars by means of the presence of a specific active site on the surface of the molecule. And this generality, in turn, is what makes E. Coli sensitive to be fooled by the same artificial sweeteners as may human beings. So, Peircean symbols are not a human prerequisite only. Another related problem is that even if the icon-index-symbol triad is oriented from the simple towards the complex, it is not compositional. Deacon’s reconstruction makes indices consist of a specific configuration of icons, and the symbol consist of a specific configuration of indices. But pure icons form a limit concept in Peirce – they will vaguely signify any possible object resembling them, because they are not connected to any actual object (that being the task of an index) – so an index could never result from any combination of such vague, dream-like signs. The pure index is also a limit category – like a push in the back or a pointing gesture directing attention to an object. Such signs are indeed possible, but they remain marginal, because neither the pure icon nor the pure index is able to communicate anything. In typical usage, indices are connected with icons in propositions, bearing information about the object which the index merely indicates without itself giving any information at all. Finally, a pure symbol without any iconical or indexical qualities are equally impossible – in order to be understood, a symbol must bear information in the shape of an icon and relate that information to an object by means of an index.

Thus, the collaboration of icons and indices within symbols is a way of describing the triad much closer to the actual functioning of signs than the focusing upon rare, detached specimens of the three aspects of sign use. This forms the basic reason why the tempting idea of mapping the icon-index-symbol triad onto the process of evolution is doomed to fail: pure icons, indices, symbols are marginal phenomena. So, there could never have been an evolutionary period where purely iconic signs prevailed – they are much too vague to communicate any information of value for biological processes, because their content is merely possible and does not relate to the actual world. And there could never have been a purely indexical period – indices being attention-directing and based on the here-and-now, they are unable to perform the central task of orienting and guiding biological activity into the future like the generality of the symbol.

I do agree, however, with Deacon in the conviction that a central part of the semiotic-missing-link issue is conceptual – and also in the contention that Peirce’s semiotics may yield some of the conceptual tools needed for approaching that issue. The whole edifice of Peircean semiotics constructed in the years after the turn of the century rather forms a physiology of the semiotic reasoning process than forming a composition of complex signs out of simpler signs. Thus, in one of the important papers laying out the principles of his semiotics, the Kaina Stoicheia (New Elements) of 1904, Peirce describes symbols as “genuine” signs, by contrast taking simpler signs (like icons and indices) to be “degenerate” signs – indices degenerate to the first degree and icons degenerate to the second degree.Footnote 1 Degeneracy here means non-generic (and hence rare) at the same time as it means incomplete, in need of completion. As examples of symbols, single words or concepts are much too often selected; this is to some degree misleading because the central symbol subtype to Peirce is not the single word but the proposition (or Dicent, or Dicisign, or Pheme, in the ever-growing wilderness of terminology which constitutes part of the reason for the ignorance of his doctrine) – the sign that claims something about something.Footnote 2 So, the working of icons and indices within the confines of Dicisigns is the key to the relation between the sign types in pragmatic reality. In his essay on symbols in the present volume, Deacon actually approaches the issue of Dicisigns when he emphasizes the importance of the combination of indices with symbolic predicates in actual sign use (even if he leaves out of consideration the explicit discussion of Peirce’s third triad – that of Rheme-Dicisign-Argument (or Term-Proposition-Argument)). By contrast I find this triad and its emphasis on reasoning is a key also to the icon-index-symbol issue.

2 The Ubiquity of Dicisigns

The wide applicability of the icon-index-symbol triad in semiotics often seems to have pushed other important aspects of Peirce’s semiotics into the background – such as his third trichotomy, forming his version of the classic logical distinction between term, proposition, and argument. Peirce uses changing terminology about this triad, the most stable seeming to be Rheme, Dicisign, Argument.Footnote 3 Rhemes comprise predicate terms understood as unsatiated propositions – like “_ is blue” or “_ gives _ to _” where the filing-in of one or more of the empty slots (indicated by underscore) by a rhematic subject will give a proposition. Peirce’s analysis of the proposition, hence, is a version of the traditional subject-predicate distinction – where each predicate may take several subjects (like 3 in the case of “give”), not unlike Frege’s analysis of the proposition in terms of function and arguments.

Dicisigns, now, are related to icons and indices by the fact that Dicisigns have two parts, subject and predicate, with indexical and iconical properties, respectively. The predicate part of the Dicisign must, directly or indirectly, convey some iconic content for the description of the propositional subject. It may work directly, by means of the presentation of a picture or a diagram, or indirectly, by means of a linguistic or other conventional predicate symbol, in turn referring to schematic iconic content. The subject part(s) of the Dicisign, on the other hand, must, directly or indirectly, bind this informational content of the rheme to some object(s) to which the Dicisign refers.Footnote 4 Again, this may take place directly, by pointing or by a rigid designator like a proper name, or more indirectly by means of a symbolic index like a pronoun or a common noun. The simple Dicisign “Alfred is happy” has a proper name as linguistic index which fills in the empty slot of the predicate “_is happy”. A royal portrait is also a Dicisign, uniting the linguistic index of “Louis XIV” on the picture frame with the image predicate provided by the painted man on the canvas. This double architecture of the Dicisign, of course, is basically motivated by it saying something (predicate) about something (subject). These two aspects of the Dicisign now point, each in their way, description and reference, respectively, to the same object. So Peirce may describe the Dicisign as a special sign uniting two different signs both related to the same object, but in two different ways.Footnote 5 The possibility of the Dicisign to possess a truth value comes from the possibility of those two aspects actually, successfully, fitting the same object. It is this general truth-bearing capability of Dicisigns which makes them central to logic and reasoning – and which makes isolated icons or indices “degenerate” by comparison to full-blown symbols.Footnote 6

These simple observations, however, make Dicisigns indispensable for biological sign use. Signs which may not convey truth are hardly efficient in biology: icons only indicating vague possibilities have little pragmatic efficiency in cognition and communication, just like isolated indices only able to indicate that something is happening but not what it is, may be of local use but not much more than that. This is why Dicisigns are ubiquitous in biology. This has been hard for both biologists and biosemioticians to appreciate, probably because of the widespread idea that propositions require the judgment of a conscious “propositional stance” found only in human beings. Here, Peirce’s purely semiotic definition of the proposition as a Dicisign combining two signs into one irreducible whole gives us a formal notion of Dicisign – neither presupposing consciousness nor explicit acts of judging. Thus, “Dicisign” in this definition is a much broader concept than explicit, linguistic “propositions” strictly; rather it corresponds to “proto-propositions” as defined by Hurford (2007) in a parallel argument for the existence of basic logical structure in animal cognition. Pragmatically, the existence of Dicisigns will be displayed by specific perception-action connections – in an organism’s behavioral possibility of acting in a typical, categorized way prompted by the categorical perception of some biologically important, stable feature of its environment. Thus, when the E. Coli reads the perimeter of the carbohydrate molecule, its subsequent oriented swimming counts as the proof that a Dicisign combining the abstract shape of the active site with a here-and-now presence of such a site has been processed by the bacterium. Of course, the molecular surface configuration of this active site may appear in other inert macromolecules without any Dicisign to be realized – its “activity” is only granted by the bacterium. The decisive precondition is that the receptor molecules of the E.Coli make it change behaviour in a characteristic and typical way, oriented towards the continuation of its metabolism (and hence its survival).

This simple biological example gives us the important clue to what is semiotically basic in biology: the stable metabolism of an organism. The single phases of the metabolism, may, of course, be described by purely chemical means, but it is the fact that these phases form a circular, self-sustaining structure which provides the basic biological prerequisite for adaptations towards sustaining this process better – and to act so as to support metabolism (by swimming in the right direction for digesting carbohydrates, in this case). This formed the basic insight in von Uexküll’s early biosemiotic notion of the functional circle of animals, binding together perception signs with action signs to form the fundamental cyclic interweaving of perceptions with guided action. Thus, the perceptual Dicisign of reading the active site on a carbohydrate molecule – a primitive version of the proposition “This is sugar” – is followed by the action Dicisign of swimming in that direction – to form an argument: “If sugar, swim in its direction. This is sugar. So, swim in its direction”. That this forms a (very primitive) argumentFootnote 7 – and not merely a cause-effect chain – is evident from the fact that the E. Coli may be fooled by artificial sweetener whose molecules possess the same molecular surface configuration as the active site in carbohydrates – but otherwise have a rather different chemistry without the easily releasable binding energy of carbohydrates.

This is not to say, of course, that this process is not underpinned by causal relations. The semiotic aspect of the process lies in the fact that the weak, local interaction makes a whole class of different stimuli give rise to the same, typical behaviour. Thus, it is the fact that the bacterium does not interact causally with the whole of the molecule (before consuming it, that is) but merely weakly interacts with a spot on its perimeter which is a precondition for its turning a semiotic and not merely causal process.Footnote 8 This argument structure binding together perception and action, of course, is close to being as primitive as it may get – and the explicit analysis of it into two distinct Dicisign phases is possible for the observer, but definitely not for the bacterium itself. It has no possibility to make any single aspect of the argument explicit nor autonomous – there are few chemical agents (besides carbohydrates, certain toxins) which the bacterium is able to categorize and react to. The automat-like character of the perception-action link testifies to its holist, yet undivided character. So the animal is not able to address the logical structure of its own perception-action chain as such, nor to substitute other perceptions or conclusions for those of sugar and toxin, or eating and fleeing, respectively. Still this basic argument structure is what makes it possible, during evolution, for higher animals to refine and spread perception-action cycles to much larger parts of their surroundings, therely enlarging their Umwelt, and, what is more, to isolate parts of the Argument as Dicisigns, and, in turn, parts of those Dicisigns as Subject Indices and Predicate Icons.

It is the fact than metabolism has an active perception-action phase – marginal in plants and funghi, central in animals – that introduces semiotics in the simple reasoning inherent in searching the environment for nutrients (and, in the E. Coli case, escaping toxins). The “reading” of carbohydrate and toxin gradients before a substantial concentration of either is present is what allows the animal the conclusion of going into the right direction for finding (resp. escaping) such concentrations.

3 Sign Action – A Process Differentiated Through Evolution

This discussion of E. Coli, then, serves to state the basic argument that biologically simples signs are not isolated icons or indices, only later to be composed into more complex signs.Footnote 9 Biologically simple signs, rather, are full-fledged perception-action arguments lacking explicit internal articulation – but bearing with them the possibility of later segmentation, articulation, autonomization, adaptation to further purposes, making it flexible, potentially loosening semiotic structure still more from its causal basis and eventually making the combination of separated Dicisigns, of predicates and subjects, possible. Parts of the metabolism may become relatively autonomous, forming organs – and parts of the perception part of metabolism may acquire their own parts, giving rise to cognitive plasticity, association learning, memory, recursivity and much more, just like the action part of metabolism may differentiate into motor limbs and tools able to support complicated action sequences, co-shaping the environment, depending in turn on this transformation.

The basic argument for this ubiquity of simple proto-propositions linking into arguments in biology is thus based on the observation that phylogenetically acquired habits – like bacteria swimming in the direction of sugar – must be both simple, stable, and true in order to support survival. If not simple, it would be beyond simple animals to process them. If not stable, they would not be able to address stable features of the environment (such as carbohydrate’s combination of easily digestible binding energy and characteristic active sites). And if not more often true than false, they would lead to the perishing of the lineage rather than its survival. And Dicisigns are signs able to express truths. This points to the fact that semiotic evolution should not be seen as going from the simple to the complex in terms of beginning with atomic signs which later serve as building blocks for more complex signs. The process from simple to complex should be conceived of in a non-compositional way: the overall semiotic argument process structure is there from the metabolic beginning – and semiotic evolution rather takes the shape of the ongoing subdivision, articulation, and sophistication of primitive signs, an ongoing refinement of parts and aspects acquiring still more autonomy. Hence, on this view, semiotic compositionality rather forms an important achievement than it forms the starting principle: the ongoing autonomization of parts and aspects of Dicisigns and their combinations may make them more and more compositional – resulting in a growth of combination possibilities and hence increasing cognitive plasticity. Such segmentation of the argument process thus constitutes the overall shape of the increase in “semiotic freedom” during evolution, highlighted by Hoffmeyer (2010). The reason for taking, once again, the textbook example of E. Coli to illustrate basic sign use, is thus to insist on the fact that the kernel of semiotic cognition is the extremely simple piece of reasoning which connects perception and action. The fact that this process may err is what proves its character of (simple) reasoning. It also indicates that cognition begins long before organisms with central nervous systems arise, and even longer before the appearance of organisms with movable perception organs, binding of different sensory inputs into cross-modal perceptions and the construction of environment mappings. This should make us cautious with more or less automatic assumptions that nervous tissue such as found in the CNS of higher animals should be the privileged locus of cognitive processes. Such an idea is a sort of magic, ascribing special abilities of wonder to neurons – overlooking that cognition is a process connecting the whole of an organism to aspects of its surroundings. And such an idea is what Clark (2008), pointing to external elements of cognitive processes, calls “neurocentrism”. Not only may “anthropocentrism” make us think that all accidental properties of human cognition are properties of cognition as such – but “neurocentrism” may repeat the error on a larger scale presuming that properties of nervous tissues are properties of cognition as such. Rather, cognition exists in the shape of perception-action cycles long before the evolution of multicellular organisms and the ensuing evolution of central nervous systems. Such evolution processes should rather be conceived of as adaptations to make the interface between perception and action more plastic, more versatile and add to the structure of dicisigns which the organism is able to process before turning to action. So, the specialization of certain cells to become neurons, interlinked in the CNS in multicellular organisms forms a way of adapting the organism not only to its specific surroundings – but to achieve still more complicated logical structures and reasoning capabilities. Neurons and CNS’s are special adaptations to the requirement of complicated cognitive and logic processes – which is why they have to adapt to necessary structures of such processes. As Peirce says – bearing in mind his objective notion of “mind”: “For we must remember that the organism has not made the mind, but is only adapted to it. It has become adapted to it by an evolutionary process so that it is not far from correct to say that it is the mind that has made the organism.” (“Abstract of 8 lectures”, undated, NEM IV, 141). The more varied the problems posed by the surroundings, including fellow conspecifics, become, the more complicated the intervening structure between perception and action must develop in order to adapt plastically – and the more that structure must conform to basic regularities of semiotics and logic: “ Logic, for me, is the study of the essential conditions to which signs must conform in order to function as such.” (Peirce, “New Elements”, 1904; Peirce 1976 IV, 235ff) And, what is more, the more variation the environment presents within the small ontogenetic time-scale window of single organism life, the more of reasoning must be transported from the slow process of phylogenetic Darwinian adaption (teaching, e.g., bacteria, over millions of generations, the Argument habit of following sugar and avoiding certain toxins) to the comparatively extremely quick process of ontogenetic adaptive learning (teaching, e.g., apes, over a period of days, to acquire the habit of associating a specific location in the jungle with the presence of specific fruits). Here, biosemioticians must learn not to commit the time-scale error of automatically taking long-term habits for being non-semiotical while short-term habits are much easier seen as having a semiotical nature.Footnote 10 The actual sign-exchange, both in the bacterium and the ape example, always takes place in the vanishing moments of individual ontogenetic lifetime – whether the underlying habit is constituted in the faster timescale of phylogenetic adaptation or in the faster timescale of ontogenetic learning does not make any principal difference (albeit an enormous difference in behavioural plasticity, it is true).

Let us sum up then, the character of primitive, metabolic argument. It connects a perception Dicisign with an action Dicisign to an Argument which, again, forms part of the overall metabolism of the cell. The reason for calling it an Argument is its ability to attain truth – and to err, respectively.Footnote 11 As an Argument, of course, it lacks a series of important aspects characterizing explicit arguments made by human beings. The connection between its part has been established over the vast phylogentic timescale of evolution and could only be changed in the same way. There is no ontogenetic freedom to exchange the premises for other premises in an online trial-and-error process. No matter which consciousness definition you adhere to, there is no reason to assume any counscious access to the conclusion or to other parts or aspects of the argument structure. The argument appears as a behavioral gestalt, whose parts are only accessible as such to the external observer and analyst, not to the bacterium.

Finally, this overall argument implies that the distinction between man and animal must be sought elsewhere than in a distinction between icons/indices on the one hand and symbols on the other – namely in the growing degree of explicit control and metasemiotics, the ability for an organism to make explicit and control its own signs.

4 Hypostatic Abstraction

With the intensified research into human prehistory occupying many different disciplines, a Pandora’s box of old questions has been reopened: the origin of language, the emergence of culture, the physical anthropology and evolution of human beings – and, conversely, the issue of communicative and cogntive abilities of other higher animals as compared to those of human beings. What is specific to human semiotic and cognitive abilities as compared to those of higher animals? A series of different answers to this issue of the semiotic or cognitive “missing link” between higher animals and human beings are already on the market. Symbol use (Terrence Deacon), joint attention (Jerome Bruner, Michael Tomasello), language syntax (Chomsky), specific types of “blending” (Fauconnier & Turner, 2002), etc. The discussion is both electrified and muddled by the fact that these hypotheses range over different fields such as psychology, linguistics, semiotics, cognitive science, etc. This implies that the proposals mentioned are not even directly comparable – in order to be compared, they should so to speak be translated into each others’ terminology. What would, e.g., the psychological notion of “joint attention” amount to if translated into the terminology of linguistics, semiotics, or neuroscience? The possibility exists that it might turn out to mean approximately or even exactly the same as one or several of the other proposals – this could only be decided after such a reconstruction process which is, by no means, a simple translation issue but a reconstruction which will only be possible after a process of conceptual and empirical development and which will, in itself, constitute a main part of a solution.

I myself have aired the idea that a good candidate for this semiotic-cognitive “missing link” might be Peirce’s notion of “hypostatic abstraction” (Stjernfelt, 2007, chapter 11). Of course, no single semiotic feature may presumably be held responsible for all semiotic and cognitive differences between human beings and higher animals – still I find hypostatic abstraction to be one of the central candidate devices because permitting the making explicit and controling of various prehuman semiotic capacities and hence indispensable for the construction of human thought and language. I shall begin by presenting Peirce’s discussions of the termFootnote 12 – even if the matter may of course not be solved by mere Peirce philology only.

Peirce never wrote a comprehensive treatise on the issue but returns to it over and over again in his mature work around the turn of the century.

Here, it refers a process as well as a product, to be found in a bundle of related semio-cognitive events:

  • Linguistically: the construction of an (abstract) noun from more concrete expressions, such as a adjective (“hard” → “hardness”), a verb (“give” → “giver”, “gift”, “given”), or a (more concrete) noun (“object” → “objecthood”), etc. From the sentence “The sky is blue”, the sentence “The sky possesses blueness” is constructed.

  • Logically: the corresponding construction of a subject on the basis of a predicate, thus adding a new 2nd-order individual (“blueness”) to the domain represented.

  • Mathematically: the application of a meta-level operation or object regulating other, more basic operations or objects (e.g., passing from the existence of different types of connections between entities to forming the concept of “relation” as a new abstract object. The properties of this object now become open to investigation in higher-level hypostatic abstractions (the “symmetry”, “transitivity”, etc. of relations may now be investigated).Footnote 13

  • Cognitively: the process of taking a thought for a thing, so that a new cognitive object is constructed on the basis of a thought – alternatively described as the “stiffening” of transient, fleeting cognitive content into a stable shape facilitating further reasoning pertaining to this new, abstract object.

  • Perceptually: the spatialization of a temporal perception process, such as forming the trajectory as an abstraction from the locations covered by the perceived movement of an object: going from “a point moves” to “the line traced by the moving point”.Footnote 14

Peirce’s ambitious ideas are that these rather different examples constitute different occurrences of the same basic cognitive-logical structure and process characterized by their result: the occurrence of a new, higher-level cognitive object. Most often, these aspects of hypostatic abstraction are merely mentioned as examples; they are not explicitly distinguished and interrelated as subtypes of the concept. The basic cognitive purposivity of hypostatic abstraction stands out as its raison-d’être: it facilitates the explicit reasoning and investigation pertainting to general issues which would otherwise remain implicit, transient or lost in concrete particulars (Of course, once hypostatic abstraction is possible it need not build on existing particulars and thus may refer to non-existing or fictive universals). The many different linguistic devices for hypostatic abstraction are tools which further develop, detail, and make explicit aspects of the ongoing cognitive process of reasoning. Thus, the adjective “red” basically refers to particular, concrete, here-and-now occurrences of that color and allows for their comparison, while the noun “redness” (or “the color red”, “the red”, etc.) constructs a new, stable, abstract object interconnecting these different occurrences and makes possible the further reasoning on this color as such, abstracted from its concrete occurrences, and on its relation to other colors, other properties etc.

Hypostatic abstraction may be described as a simple deduction from a premise “This object is red” to a conclusion “Redness exists (in this object)”, so that it makes sense to say that the hypostatic abstraction is an entity whose being consists in the (purported) truth of a predicate expression:

For by means of abstraction the transitory elements of thought, the {epea pteroenta}, are made substantive elements, as James terms them, {epea apteroenta}. It thus becomes possible to study their relations and to apply to these relations discoveries already made respecting analogous relations. In this way, for example, operations become themselves the subjects of operations.

To take a most elementary example -- from the idea of a particle moving, we pass to the idea of a particle describing a line. This line is then thought as moving, and so as generating a surface; and so the relations of surfaces become the subject of thought. An abstraction is an ens rationis whose being consists in the truth of an ordinary predication. (“Relatives” in Baldwin’s Dictionary, 1901; CP 3.642)

This should not be taken to imply that hypostatic abstraction expressions referring to non-existing objects may not exist. “Unicornicity” is a hypostatic abstraction from “unicorn” even if no unicorns exist – the implicaiton of Peirce’s definition is that, in this case, the hypostatic abstraction does not have any “being”, that is, it does not refer to any real possibility like in the cases of “redness” or “hardness” or “trajectory”. The deductive character of hypostatic abstraction is not changed by this observation – like any deduction, its validity dependes on the soundness of the premiss invoked: “If and only if x exists, then x-ness has being”. Thus, if no x exists, x-ness has no being, – but in many cases the validity of this claim may be investigated both by investigating x’s and investigating x-ness. Thus, the hypostatic abstractive deduction forms no guarantee that the resulting abstraction has a fundamentum in re and refers to really existing kinds – the well-known examples of fallacious hypostatic abstractions in science such as “phlogiston” or “caloric” testify to that.

The fact that the hypostatic abstraction is a deduction has often been confused with the possibly abductive character of the reasoning process in which a hypostatic abstraction may take part. Making a hypostatic abstraction may, in many cases, be part of a trial-and-error reasoning process where the abstraction made is subject to further investigation so as to determine its degree of reality. Thus, the deductive step of hypostatic abstraction forms, in this broader perspecive, part of an abduction whose valitidy must be investigated by further de- or induction on the basis of the abstraction made. This is why hypostatic abstraction has sometimes been characterized as deduction, sometimes as abduction (cf. Pape and Short in Houser, Roberts, & van Evra, 1997; cf. Stjernfelt, 2007, p. 458).

It should immediately be added that Peirce takes great care to distinguish this process from what he calls “distinction”, the attention ability which permits the focusing on a particular part or aspect of an object at the expense of other parts or aspects of that object – and which is often confounded with hypostatic abstraction. These focusing abilities come in three variants, nicknamed “dissociation”, “prescission”, and “discrimination”, respectively. Dissociation is what permits the distinction between different independent qualities, such as “red” from “blue”: prescission is what permits the distinction of a part which may be supposed to exist independently of another part, such as “space” from “color”, while discrimination is what permits the distinction of a part which may be only imagined separately, such as “color” from “space”. These two latter distinction types are important to the investigation of objects involving features dependent on each other in different patternings. The kind of attention they pertain to, however, involves imagining the object endowed with indeterminate parts:

In general, prescission is always accomplished by imagining ourselves in situations in which certain elements of fact cannot be ascertained. This is a different and more complicated operation than merely attending to one element and neglecting the rest. (CP 2.428).

According to Peirce, it is of paramount importance to keep the distinctions apart from hypostatic abstraction – while the former pertain to the degree of particularity and generality and thus permit chains of increasing generality like “red” – “color” – “hue”, the latter does not lead to higher generality but to the creation of new, abstract or ideal objects of thought or discourse.Footnote 15 The distinction between the two may be expressed as follows:

But even in the very first passage in which abstraction occurs as a term of logic, two distinct meanings of it are given, the one the contemplation of a form apart from matter, as when we think of whiteness, and the other the thinking of a nature indifferenter, or without regard to the differences of its individuals, as when we think of a white thing, generally. The latter process is called, also, precision (or better, prescission): and it would greatly contribute to perspicuity of thought and expression if we were to return to the usage of the best scholastic doctors and designate it by that name exclusively, restricting abstraction to the former process by which we obtain notions corresponding to the “abstract nouns.” (CP 2.427, “Terminology” 1893).

In most concrete cases, of course, the two procedures work closely in tandem: before the hypostatic abstraction of “redness”, a distinction is required to isolate the property of “red” in the object (more generally, hypostatic abstraction is impossible without a preceding distinction) – but still, the working of the two must be kept analytically distinct. In contrast to many empiricist theories of abstraction, moreover, it should be added that Peirce does not identify any of the two with induction as the statistical investigation of properties in a sample of objects. Abstraction does not presuppose induction and it is perfectly possible to perform a hypostatic abstraction on the basis of one observed object only (even if it may be wiser to perform it after an induction summing up knowledge of a wider sample of objects).

An important aspect of hypostatic abstraction is that, in making a second-order object out of a thought, it gives it concrete form and thus facilitates cognitive and logic manipulation and investigation of it – as if it were a particular individual object:

Intuition is the regarding of the abstract in a concrete form, by the realistic hypostatization of relations; that is the one sole method of valuable thought. Very shallow is the prevalent notion that this is something to be avoided. You might as well say at once that reasoning is to be avoided because it has led to so much error; quite in the same philistine line of thought would that be; and so well in accord with the spirit of nominalism that I wonder some one does not put it forward. The true precept is not to abstain from hypostatization, but to do it intelligently. … (CP 3.383, “A Guess at the Riddle”, c. 1890).

This implies that the hypostatically abstract object may be seen as if it shared some of the characteristics of particular individuals: it has properties, it stands in various relations to other such objects, it may be subsumed by still higher genera – in that sense hypostatic abstraction is a simplifying device involving cognitive economy because it permits to use some of the same means for their investigation which we use interacting with particulars. Peirce also ascribes abstractions a seminal role in his famous distinction between corollarial and theorematic deductions,Footnote 16 the former only relying upon definition of concepts appearing in the premises, the latter requiring the introduction of additional elements in the shape of postulates to conduct the proof. Theorematical reasoning, of course, requires creativity and guessing, even if being deductive – and the most challenging theorematical deductions are taken to involve the introduction of abstractions:

“Deductions are of two kinds, which I call corollarial and theorematic. The corollarial are those reasonings by which all corollaries and the majority of what are called theorems are deduced; the theorematic are those by which the major theorems are deduced. If you take the thesis of a corollary, – i.e. the proposition to be proved, and carefully analyze its meaning, by substituting for each term its definition, you will find that its truth follows, in a straightforward manner, from previous propositions similarly analyzed. But when it comes to proving a major theorem, you will very often find you have need of a lemma, which is a demonstrable proposition about something outside the subject of inquiry; and even if a lemma does not have to be demonstrated, it is necessary to introduce the definition of something which the thesis of the theorem does not contemplate. In the most remarkable cases, this is some abstraction; that is to say, a subject whose existence consists in some fact about other things. Such, for example, are operations considered as in themselves subject to operation; lines, which are nothing but descriptions of the motion of a particle, considered as being themselves movable; collections; numbers; and the like. When the reform of mathematical reasoning now going on is complete, it will be seen that every such supposition ought to be supported by a proper postulate. At any rate Kant himself ought to admit, and would admit if he were alive today, that the conclusion of reasoning of this kind, although it is strictly deductive, does not flow from definitions alone, but that postulates are requisite for it.” (CP 7.204; “On the Logic of drawing History from Ancient Documents especially from Testimonies”, 1901).

As to the discussion of the semiotic “missing link” it should be mentioned that many higher animals are able to make prescissions – the ability to isolate features in an object is the precondition for associative learning, linking up co-occuring such features – just like they are able to make deductions on the basis of phylogenetically inherited or ontogenetically acquired habits. Theorematic reasoning and Hypostatic abstractions, on the other hand, seem to be missing among animal proto-concepts.

5 Self-Control by Abstraction in Human Semiotics

In a central argument, Peirce links the special semiotic and cognitive abilities in human beings to a higher degree of self-control which is, in turn, connected to the ability to make hypostatic abstractions. Let us first scrutinize his notion of “self-control”. An important idea here is that self-control is crucial for inferences to count as real reasonings, as he epigrammatically may say: “... reasoning is thought subjected to self-control ...” (CP 5.533 “Pragmaticism, Prag. [4]” c. 1905). This is why computers (“logical machines”) are not taken to be able to reason – even if their actions may formally realize inference structures and they are able to produce outputs which are interpretable as truths – they do not possess any self-control. The potentiality of specific action is sufficient to count as a habit – but belief requires the self-control of habit: “[Readiness] to act in a certain way under given circumstances and when actuated by a given motive is a habit; and a deliberate, or self-controlled, habit is precisely a belief.” (5.480).

Fully realized self-control, on the other hand, may have as its result the formation of mechanical-like thought habits: “The power of self-control is certainly not a power over what one is doing at the very instant the operation of self-control is commenced. It consists (to mention only the leading constituents) first, in comparing one’s past deeds with standards, second, in rational deliberation concerning how one will act in the future, in itself a highly complicated operation, third, in the formation of a resolve, fourth, in the creation, on the basis of the resolve, of a strong determination, or modification of habit. This operation of self-control is a process in which logical sequence is converted into mechanical sequence or something of the sort. How this happens, we are in my opinion as yet entirely ignorant. There is a class of signs in which the logical sequence is at the same time a mechanical sequence and very likely this fact enters into the explanation.” (8.320. letter to F.C.S.Schiller, undated).

This, however, is only possible as the result of a complex process involving standards, future acts, and a decision to modifiy thought habits. Such self-control is required for full mastering of reasoning – but is the result of a process with simpler biological antecedents. Importantly, Peirce sees a decisive aspect of self-control in the psychological ability to isolate a thought from other intrusions – a psychological equivalent to the logical notion of “distinction” discussed above:

Contemplation consists in using our self-control to remove us from the forcible intrusion of other thoughts, and in considering the interesting bearings of what may lie hidden in the icon, so as to cause the subjective intensity of it to increase. (7.555)

The isolation of the iconic sign may count as a first step in human self-control – to be followed by the hypostatic taking of that sign to be a thing in itself. Hypostatization, now, is crucially connected to the particularity of human reasoning. Peirce himself only rarely discusses hypostatic abstraction in connection to the man-animal issue. The most important locus is the following quote:

To return to self-control, which I can but slightly sketch, at this time, of course there are inhibitions and coördinations that entirely escape consciousness. There are, in the next place, modes of self-control which seem quite instinctive. Next, there is a kind of self-control which results from training. Next, a man can be his own training-master and thus control his self-control. When this point is reached much or all the training may be conducted in imagination. When a man trains himself, thus controlling control, he must have some moral rule in view, however special and irrational it may be. But next he may undertake to improve this rule; that is, to exercise a control over his control of control. To do this he must have in view something higher than an irrational rule. He must have some sort of moral principle. This, in turn, may be controlled by reference to an esthetic ideal of what is fine. There are certainly more grades than I have enumerated. Perhaps their number is indefinite. The brutes are certainly capable of more than one grade of control; but it seems to me that our superiority to them is more due to our greater number of grades of self-control than it is to our versatility.

Doctor Y. Is it not due to our faculty of language?

Pragmaticist. To my thinking that faculty is itself a phenomenon of self-control. For thinking is a kind of conduct, and is itself controllable, as everybody knows. Now the intellectual control of thinking takes place by thinking about thought. All thinking is by signs; and the brutes use signs. But they perhaps rarely think of them as signs. To do so is manifestly a second step in the use of language. Brutes use language, and seem to exercise some little control over it. But they certainly do not carry this control to anything like the same grade that we do. They do not criticize their thought logically. One extremely important grade of thinking about thought, which my logical analyses have shown to be one of chief, if not the chief, explanation of the power of mathematical reasoning, is a stock topic of ridicule among the wits. This operation is performed when something, that one has thought about any subject, is itself made a subject of thought. (...)Footnote 17 (“Pragmaticism, Prag. [4]” c. 1905, CP 5.533).

The quote gives two important arguments. The first is that self-control comes in many grades and increase during evolution. Our hypothesis here will be that this increase corresponds to the increase of articulation and segmentation of the perception-action chain into detailed argument structures. Any autonomization of a part of that chain corresponds to an increase in self-control. On top of such grades of self-control which is already present in higher animals, Peirce presents an architecture of additional human self-control grades: (1) training (2) self-training, controlling one’s own self-control, involving imagination (3) adoption of a rule guiding this meta-control, (4) improvement of that rule after some higher ethical standard, thus controlling the control over one’s control (5) controlling, in turn, that rule after some aesthetic standard (Peirce’s notion of aesthetics pertaining to all goals which are worth pursuingFootnote 18). Every such step, of course, takes the former step as its object, thus creating a newhypostatic abstraction subject to variation and evaluation.Footnote 19 Many higher animals, it is well-known, may be subject to training, but the next, decisive step of self-training seems only rudimentarily accessible to higher animals.

The other crucial argument here is that such self-control is seminal to human thought and language – and that this self-control is facilitated by thinking of our signs as signs, by thinking about thought and thereby becoming able to criticize our own thought logically. Self-control involves the taking one’s own thought as the object of a meta-level thought. But this is only possible by making the first thought an object – stiffening in the shape of a hypostatic abstraction. Such self-control even makes possible language. How should this be interpreted? – it is well known that natural language learning does not take place by the explicit memorizing of linguistic rules and that practicing knowledge of grammar does not entail any explicit insight in grammatical principles (much like Peirce’s logical distinction between implicit logica utens and explicit logica docens which differ in that the former is interested in the result of reasoning, not the process, the latter vice versa). The work performed by self-control here is more basic – it is the ability to wonder and check whether a particular sign is suitably used, focusing upon the relation between sign, object, and interpretant, upon its relation to other signs and their objects and interpretants. Such ability is taken to be the prerequisite to the establishment of grammar, fine-grained taxonomies, tuning of schematic content, expression-content couplings, etc. in the development of languages.Footnote 20

It is important to Peirce’s notion of self-control, now, that such self-control is a merely restrictive measure, selecting valuable inferences among less valuable inferences – thus, it presupposes the existence of inferences which it then, subsequently, turns into reasoning by controlling them:

But self-control is the character which distinguishes reasonings from the processes by which perceptual judgments are formed, and self-control of any kind is purely inhibitory. It originates nothing. Therefore it cannot be in the act of adoption of an inference, in the pronouncing of it to be reasonable,

that the formal conceptions in question can first emerge. It must be in the first perceiving that so one might conceivably reason. And what is the nature of that? I see that I have instinctively described the phenomenon as a “perceiving.” I do not wish to argue from words; but a word may furnish a valuable suggestion. What can our first acquaintance with an inference, when it is not yet adopted, be but a perception of the world of ideas? (5.194 Lectures on Pragmatism 1903).

The question of the roots of inference is here answered phenomenologically – investigating the origin of inference structures as seen from the perspective of human mind. Before subjecting an inference to control and evaluating it, we must be able to perceive it in “the world of ideas”. This idealist wording of course leaves out the fact that we access that world not by any direct, mystic, purely intuitive route, but only by the intermediary of diagrams, facilitated by imagination. The inhibitory work performed by the different levels of self-control, then, presupposes a wealth of possible inferences and abstract objects to chose between. The imaginative creation, variation and combination of such inferences and objects – at each of the control levels – is thus the prerequisite for inhibitory self-control to perform its function.

And posed as an evolutionary question, the basic pool of such inference structures is found in the perception-action habits refined through the evolution of animals – habits which have been subjected to increasing degrees of control already over the course of evolution, before they are made, in turn, the object of the vastly increasing human processes of self-control by means of hypostatic abstraction and diagram experimentation.Footnote 21

It is an important corollary of self-control, as described here, that it always involves (at least) two levels, that of inhibitory controlling, and that of imaginary creativity (being controlled). This implies that the focus of control must alternate between the levels, evolving inferences on the lower level and pruning them on the higher level. This makes self-control a special case of Peirce’s important idea of the dialogicity of logic. This, I think, makes it possible to compare the Peircean notion of self-control by hypostatic abstraction with Tomasello’s well-known ideas of joint attention (Tomasello, 1999, 2008).

6 Hypostatic Abstraction and Joint Attention

For how does Peirce’s idea of hypostatic abstraction fit the Tomasellian idea of joint attention? For a first glance, the two ideas may seem wide apart, but for a closer glance important similarities appear. Joint attention can not, of course, be reduced to two parties both of them intending the same object. It also involves the knowledge in each part of the other part’s attention. But even that is not sufficient. As Kaplan and Hafner insist from the point of view of implementing joint attention in robot research, joint attention is not achieved even by robots tracking other robots’ attention and coordinating that with their own attention (which is robotologically possible but does not entail joint attention). Rather, joint attention is a collaborative process, in which

… the agent must understand, monitor and direct the attentional behavior of the other agent. Joint attention can only be reached if both agents are aware of this coordination of “perspectives” towards the world. (Kaplan and Hafner 2006).

Joint attention thus requires for each agent to assume the famous “intentional stance” towards the other: the attention direction detected in the other agent is interpreted as a sign that an intention is directing that attention to some goal. But even that is not sufficient: each agent must be able to influence upon the other’s attention, for instance by directing it by means of gesture, eye movements, linguistic cues, etc. And such influence is only possible based on a skill of social interaction: the agents must be able to master turn-taking, role-switching and ritualized games, as Kaplan underlines. If no turn-taking schema is active, the agents will not know who is directing whose attention at any given moment. Thus, the apparently simple phenomenon of “joint attention” entails a whole series of interrelated concepts – a molecule of social interaction. But something similar is the case with Peircean hypostatic abstraction. It forms, of course, one of the major techniques of letting “symbols grow”, Peirce’s brief version of the Enlightenment ideal of common, increasing knowledge construction. It does that by means of its ability to take other signs as its object, thereby making their content and role explicit, and hence the possible object of scrutinizing, comparing and controlling. And this whole process of thought, according to Peirce, has an irreducible dialogic structure:

Accordingly, it is not merely a fact of human Psychology, but a necessity of Logic, that every logical evolution of thought should be dialogic. (“Prolegomena”, 1906, CP 4.551).

This necessity lies in the articulation of logic in signs as the means for communicating them from a person in one moment to the same person in the next moment:

All thinking is dialogic in form. Your self of one instant appeals to your deeper self for his assent. Consequently, all thinking is conducted in signs that are mainly of the same general structure as words; those which are not so, being of the nature of those signs of which we have need now and then in our converse with one another to eke out the defects of words, or symbols. These non-symbolic thought-signs are of two classes: first, pictures or diagrams or other images (I call them Icons) such as have to be used to explain the significations of words; and secondly, signs more or less analogous to symptoms (I call them Indices) of which the collateral observations, by which we know what a man is talking about, are examples. The Icons chiefly illustrate the significations of predicate-thoughts, the Indices the denotations of subject-thoughts. The substance of thoughts consists of these three species of ingredients. (6.338, Amazing Mazes 4, 1909).

And this gives rise to the possibility of performing logical arguments and proofs, the same person occupying alternately pro- and con-positions in an ongoing dialogic process. These important ideas have often been overlooked because Peirce’s ideas on this are only scarcely represented in his published work and have not been much discussed in the Peirce literature, but it has been highlighted in the wake of the tradition of Hintikkan game-theoretical semantics (Jaakko Hintikka, Risto Hilpinen; most recently, Pietarinen 2006 has furthered the investigation of this issue).Footnote 22 This idea occurs in the context of Peirce’s logic representation systems known as Existential Graphs whose Alpha and Beta parts are isomorphic to propositional logic and first order predicate logic with identity, respectively. Peirce’s idea is that these representations reveal a dialogic structure inherent in logical arguments. One agent, the so-called Grapheus, is responsible for the construction of the discursive world, while the other, the so-called Graphist, is responsible for counter-arguing the single steps of its construction. The two agents thus collaborate in critically investigating a logical issue and take it to conclusion, and they may, of course, often be instantiated in one and the same mind during soliloquious thought processes. In Pietarinen’s Hintikkan interpretation, they may be seen as playing a semantic game against each other, and the existence of a winning strategy on the part of one of them is the game-theoretical equivalent to the truth of that part’s argument. We shall not here go deeply into the specific means the two agents use when interacting in Peirce’s elaboration of the existential graphs, but in our context, Peirce has some important general developments of what is involved:

Now nothing can be controlled that cannot be observed while it is in action. It is therefore requisite that both minds but especially the Graphist-mind should have a power of self-observation. Moreover, control supposes a capacity in that which is to be controlled of acting in accordance with definite general tendencies of a tolerable stable nature, which implies a reality in this governing principle. But these habits, so to call them, must be capable of being modified according to some ideal in the mind of the controlling agent; and this controlling agent is to be the very same as the agent controlled; the control extending even to the modes of control themselves, since we suppose that the interpreter-mind under the guidance of the Graphist-mind discusses the rationale of logic itself. (MS 280: 30–32, quoted from Pietarinen 2006).

The dialogic structure facilitates control of the thought process, because one part’s utterance in the game takes the other part’s utterance as its object in a hypostatic abstraction.Footnote 23 What is visible, of course, is only the other part’s manifest utterance, but that utterance is the response to the whole preceding game and, in that respect, indirectly refers to it – much like a move in a chess game implicitly refers to the whole preceding game and one player’s interpretation of the other’s intention as perceived from his move sequence. Of course, hypostatic abstractions are not possible within the representation systems offered in the Alpha and Beta parts of the Existential Graphs (hypostatic abstractions quantify over other signs such as precidates and thus belong to second-order logic). Peirce envisaged this second order part in his Gamma graphs which were to comprise a part aimed at the explicit representation of hypostatic abstractions. But naturally, this representation of them takes place in a hypostatic abstraction of second order.

If logical thinking necessarily possesses a dialogic structure, it forces the individual engaging in such thinking to divide so as to accomodate to it:

There is no reason why “thought,” in what has just been said, should be taken in that narrow sense in which silence and darkness are favorable to thought. It should rather be understood as covering all rational life, so that an experiment shall be an operation of thought. Of course, that ultimate state of habit to which the action of self-control ultimately tends, where no room is left for further self-control, is, in the case of thought, the state of fixed belief, or perfect knowledge.

Two things here are all-important to assure oneself of and to remember. The first is that a person is not absolutely an individual. His thoughts are what he is “saying to himself,” that is, is saying to that other self that is just coming into life in the flow of time. When one reasons, it is that critical self that one is trying to persuade; and all thought whatsoever is a sign, and is mostly of the nature of language. The second thing to remember is that the man’s circle of society (however widely or narrowly this phrase may be understood), is a sort of loosely compacted person, in some respects of higher rank than the person of an individual organism. It is these two things alone that render it possible for you – but only in the abstract, and in a Pickwickian sense – to distinguish between absolute truth and what you do not doubt. (5.420–421, “What Pragmatism Is”, 1905).

If this is the case, there seems to be a deep connection between the dialogic structure of reasoning and self-control on the level of semiotics and logic – and the central place enjoyed by joint attention on the level of human psychology according to Tomasello’s hypothesis. The parent-child dyad interaction trains the child in the first human level of self-control with he parent as the teacher, of course, but with the continuous exchange of positions making it possible for the child to experience the dialogue structure and internalize it for the benefit of its critical self-control abilities (and for later social interactions as well, of course).

Thus, a hypothesis can be stated that there is a connection between the human ability to use signs about signs and thereby exercize semiotic self-control, on the one hand, and the ability of human beings to engage in joint attentions with other subjects shaping a shared world informed by shared thoughts in the shape of shared diagram experiments.

The overall argument of this chapter, then, is that a the gradual appearance of logic and semiotic capabilities during evolution forms the backbone of the increase of cognitive competences from simple biology to higher animals and human beings. This appearance takes the shape of the ongoing articulation, subdivision and making explicit of a basic argument structure inherent in perception-action loops. The basic reason is that biological semiosis must be oriented toward adequacy truths for survival reasons, making biological cognition acutely dependent upon the ability to perceive and act in a way which is adequate to the environment, thus expressing the linking of proto-propositions. This is not to say that issues like the emergence of communication, awareness, consciousness, emotions, episodic memory, human language and much else are not important. Quite on the contrary, in this framework, such capabilities arise during evolution in order to enhance, speed up, widen, and control the basic, biological process of argumentative cognition.