Bad Grand-schema Philosophy of Information and where to find it (China)

Updated: Apr 23

[Part 2 of a 2-post special]

I’ve been investigating the sad state of grand-schematic and unificatory philosophy of information in The West, and as promised in my previous post, it’s time to turn our attention to The Middle Kingdom. Unfortunately, our tale of relative woe is international (it’s not all bad news – to be sure – but there’s lot of bad news). China, where philosophers of information are currently multiplying and on the ascent, is currently not a place to find good philosophy of information either. Not good grand schema philosophy of information, at least. When it comes to the philosophy of information, Chinese philosophers often tend to regurgitate and re-interpret Western philosophers who are either esoteric, or whose philosophies and ontologies of information are very problematic and exhibit serious flaws, inconsistencies, ontic inflation, awkward marrying of disparate concepts, and outright conceptual and explanatory contradictions. Alternatively, or often additionally, Chinese scholars attempt syntheses and unification with existing traditional Chinese and even Chinese-Marxist materialist and scientistic metaphysics. This exacerbates a-priorist pluralism-driven outcomes (Liqian & Brier, 2015).

I favour scientific metaphysics that either minimises or completely elides a-priori conceptual analysis. Chinese philosophers are (unsurprisingly) commonly naturalistic and pro-science too. Nevertheless, attempts at grand-schema unifications by Chinese philosophers of information with a scientistic and naturalist leaning are frequently just a combination of (largely random) conceptual cherry picking and difficult or baroque attempts at synthesis that would be better left untried, and that ironically fail to deliver any conclusive unification of any kind (or even a coherent notion of the nature of information).

This is all perhaps hardly surprising, due to the academic culture of Chinese scholars and the practices of the Western philosophers whom they tend to follow due to epistemic humility and a penchant for fluid conceptual unification. That said, problems equally arise due to a very strong culturally based predilection for (at least claiming) originality among Chinese philosophers of information. In some cases – there’s a palpable expression of an understandable will to compete and dominate the theoretical landscape. Unfortunately, this often drives the machine (mechanisms?) of otiose and redundant conceptual pluralism, ontological inflation, and theoretical fecundity even more.

Chinese academic culture, like that in the West, is just as much (perhaps more) about relationships and academic social networking than it is about the qualities of the work of a given scholar. I have personally known at least two Chinese philosophers of information whose work involves essentially two steps:

1. Become familiar with the work of one Western philosopher of information (and it seems better if that philosopher is not mainstream or well received in the West) that they have met at a conference or on exchange.

2. Attempt to promote that philosopher’s work, and plug it into a framework of either liberal naturalist or esoteric Chinese philosophy.

[Philosophy may not be a discipline that seems a good home for merchants and evangelists, but of course philosophers themselves know better. Imagine the book sales if one can get one’s ideas to take off in The Middle Kingdom. It’s better if you know a Chinese philosopher who is bilingual to help with those pesky translations, of course. Thar’s gold in them thar’ hills and mercantile tradin’ to be done.]

There are exceptions. As we’ll find, however, they usually fall afoul of other grand-schematic and inappropriate a-priorist commitments. While there is plenty of latitude in philosophy for thinkers to produce almost any theory within reason (usually), philosophers who are working in interdisciplinary settings involving the philosophy of science and science (a favoured approach of most contemporary philosophers of information in both the East and West) need to tread more carefully. Chinese philosophers of information of Wu’s era have the potential advantage of a background in scientifically inspired naturalistic materialism in keeping with Marxist and Maoist thought – the dialectics of nature that were part of dialectical materialism in the set of Stalin textbooks authoritative in China in the mid 20th century (an advantage from the perspective of liberal naturalism, at least) (Liqian & Brier, 2015, pp. 49–53). However, in most cases they manage to turn this advantage into an a-priorist adventure in ontological inflation and unnecessary theoretic commitments and conceptual ‘kludges’, just like their Western counterparts (whom they are often following ‘down the rabbit hole’).

The leading Chinese philosopher of information – and the most recognised domestically and internationally – is Professor Wu Kun. Wu is famous in China for his program of combining (with heavy revision) old Marxist naturalistic philosophical and pseudo-scientific doctrines (like dialectical materialism) with a metaphysical outlook that is intended to be both highly original and to accommodate mainstream scientific advances in such fields as systems theory, cybernetics, evolutionary biology, physics, and genetics.

It’s my intention to broadly criticise Chinese philosophy of information that is grand-schematic (although no more than I do for Western philosophy of information) and so I will allow Chinese philosophers of information themselves to do some of the work for me. Here is an observation by a philosophical opponent of Wu, Zhou Liqian, whose predilections are subjectivist and whose affections are correspondingly for the absential modal conception of information content developed by Terrence Deacon:

Wu’s philosophy of information represents a search for a proper transdisciplinary framework of information covering objective laws, subjective meaning and intersubjective normativity on an extended view of dialectical materialism. Unlike other philosophers of science and technology, Wu thinks that what the theory of information brings to philosophy is not additive: It’s not just that information science and studies brings new philosophical problems to the fore, or that traditional philosophy can fruitfully be inspired by information theories. It is Wu’s opinion that the discovery of the information concept calls for an absolutely new meta-philosophy or first philosophy. The discovery of information as a basic concept is such a fundamental turn of philosophy that every philosophical problem should be reconsidered in the light of this fundamental change. All parts of philosophy, including ontology, epistemology, methodology and axiology, should be rebuilt on the basis of this new philosophy of information. (Liqian & Brier, 2015)

For the uninitiated – meta-philosophy plus first philosophy is grand schema philosophy.

I recently reviewed a paper by Wu and one of his postdoctoral students. Most notable about it was the failure to engage with any important existing recent Western philosophical investigations in scientific realism (Since Karl Popper, really) – especially ontic structural realism and the various strains of competing instrumentalism. It made no reference to ontic or epistemic structural realisms, and precious little reference to other mainstream and important Western philosophy of information or science of the mid to late 20th century. Notably absent was any mention of philosophers of information like Fred Dretske, Jerry Fodor, Ruth Millikan, Daniel Dennett, Nicholas Shea, Peter-Godfrey Smith, Otavio Bueno, or Luciano Floridi. Wu and Nan made some attempt to correct the dearth in references to structural realism based upon the review, but unfortunately the result is not stellar.

[For the record – given the common features of Wu and Deacon’s systems-theory inspired philosophies, there’s not a single reference to Wu in Deacon’s reference list or index of Incomplete Nature, either (Deacon, 2013)]

One must account for the serious difficulties with the language gap. Wu does not speak English – only Russian and Mandarin – and his co-author Nan Qiong, though a capable scholar with good English, is a novice who struggles to bridge the divide between Wu’s rand-schematic commitments and those of numerous competing grand-schematic and aspiringly ontologically unifying theories in the West and in China. However, as one can see from the image below (I though this easier and fairer than copy and reformat) reading the material is a trudge, to say the least, and the trudge-i-ness is not just because of the translation challenges (which are significant):

If the paper is intended to deliver anything like a unifying conception of information (it is), then it must be said to largely fail. In fact – it doesn’t really seem to even try. The most one gets is a description in general systems-theoretic terms of what information (which is ambiguously defined at best, and not really defined at all) seems to be doing in various different settings (general dynamical systems, complex systems, chaotic systems, feedback systems, and organic evolved systems) using some general terms of reference about structure, entropy, organisation, order, and openness. If there is something to cohere (there may be) with respect to the nature of information, it hasn’t yet cohered in Wu’s schema. What information is: that’s never made clear. At least four different statements of the nature of information are referred to (notably that of Harken and one from Aleskovskii, (2002)). All we get from Wu and Kun is that:

“Information is a philosophical category indicating indirect being. It is the self- manifestation of the existing mode and status of matter (direct being)”. (Wu & Nan, 2019, p. 1)

Information is a philosophical category? Things are not going well. It’s matter being matter in some mode? But later the paper says information is not matter. It’s a dual definition anyway. Oh dear. We even seem to be calling in parenthesised artillery from Heidegger (in fact it's Lenin)? Oh no.

Wu and Nan have a habit of making long shopping lists of theories and their general commitments, and then making clusters of general hand-waving allusions to their providing support for various premises and stipulations. There are precious few actual careful arguments or analyses (the sections on complexity, self-organisation, and chaos theory are not bad). There’s just lot’s of systems theoretic stipulation and just-so storytelling (although some of it is picturesque and demonstrates a good grasp of systems theoretic precepts across disciplines and applications).

By the time one gets to the second sentence of the introduction, for example, it’s not clear what has been revealed about self-organisation: that there are lots of theories about it, that it’s meant to be true, that many self-organising systems have been discovered, or that there is just a lot of work in the field (or something else altogether). The second sentence introduces the argument-by-stipulation approach of the work (one of my favourites). It’s one thing to have a scientific-metaphysical methodological and metaphilosophical commitment – which the authors clearly have. However, making a convincing case for a grand-schematic metaphysics based upon scientific and positive scientistic premises (optimistic meta-induction, for example) is not just a matter of waving one’s hand at a large set of (albeit related) theories that seem to use the same concepts and have similar foundations and some apparent terminological and conceptual overlap.

The paper is replete with references to open systems (and system openness), system structure, entropy, systems theory (there was already some reference to this in the older version) and self-organisation. This is testament to Wu sharing with Deacon an overarching commitment to systems theory, but the interest in the centrality of autopoiesis and self-organisation in systems is newer to Wu’s material. There is now a section on the concept of hetero-organisation in the paper that matches and echoes (using different terminology) Terrence Deacon’s concept and characterisation of Boltzmann energy as external work applied to a system:

The orderly structure of self-organization can only be formed and maintained by absorbing material (mass), energy, and information from the outside world. (Wu & Nan, 2019, p. 4)

Wu and Nan have developed a theory that is basically exactly parallel to that of Deacon (a relabelled copy, in fact), but they have changed the terms (and avoided Deacon’s practice of cascading of neologisms). See especially the end of section 3.4 and the beginning of section 3.5 for a paraphrasing of what is essentially Deacon’s view of recursively self-reinforcing interpretants (Deacon, 2008, pp. 184, 189, 190-2). Along the way they make some basically wrong, or else convoluted and obscure, statements that are hard to attribute to the language gap. Here are a few examples:

A pattern is essentially a relationship, an architecture, a way of organization, a structure of order, and such things as a relationship, an architecture, a way of organization, and a structure of order can only be explained by information activities. (Wu & Nan, 2019, p. 1)

Few philosophers would agree that a pattern is a relationship. Structure may be relationships plural, depending upon one’s commitment to ontic structural realism and to the ontic priority of relations in structure, for example (Arenhart & Bueno, 2014; Berenstain & Ladyman, 2012; Esfeld, 2017; Floridi, 2008; Psillos, 2006; Wang, 2008). The idea that structure can only be explained by information activities is just wrong. Even if ‘information activities’ means information dynamics, and even if ‘only’ is meant to be ’best’, it’s wrong (mathematicians have been doing it quite effectively for quite some time in blissful ignorance of any informational ontological issues). At best, the relationship-structure ontic priority problem (objects are also involved) is a very open question with many competing theories. The same applies if a structure of order is identical to a pattern, if only because pattern and structure are largely synonymously deployed, or even terms that are usually interchangeable salva veritatae in scientific settings. There are strong arguments against Luciano Floridi’s idea that information is the ontic bottom of the world (Adriaans, 2010; Floridi, 2008; French & Ladyman, 2010; Long, 2018). It’s just as likely that physical structure is ontologically prior to all information at some level of reduction and abstraction. The ineliminability of the concept of, and reference to, in re ontic structure in all scientific discourse is strongly suggestive of this (See my book Scientific Metaphysics and Information – forthcoming).

Wu and Nan’s unificatory grand-schematic and a-priorist commitments make an entrance by way of overtures to the effect that the various sciences of complex and self-organising systems do not speak about information in a unified or consistent way, which is true (and it’s the basis of pluralism about information). However, then the methodological, conceptual, and metaphysical trouble really starts:

The philosophy of information has re-examined and explained the basic characteristics of self-organization process from the perspective of information activities so that we can grasp the essence of self organization phenomenon more deeply. (Wu & Nan, 2019, p. 1)

It’s re-examined them (and re-re-examined them) but certainly hasn’t explained them. That’s a claim that could only be true if there was, at minimum, a broadly scientifically and philosophically ratified conception and explanation of the nature of information, and even then there would be some work to do (and you’d still have even more work to do arguing with expressivists and pragmatists about what information is). There isn’t a ratified ur-concept. Except, of course, like Deacon, Floridi, Wolfgang Hofkirchner, Carlo Rovelli, and Fred Dretske (among others) that Wu is going to give it to us. In fact, astonishingly, he already has. Or – hasn’t. Apparently it’s never been written down, as such. Only general methodological allusions to it like the one already mentioned that makes information a philosophical category of mody massy something or other, where mass is Lenin’s kind of mass. Wu’s grand and unifying ontology is different to all the rest, but it’s authoritative:

The philosophy of information created by Professor Wu establishes the ontological status of information on the ontological level of philosophy on the basis of combining the relevant modern scientific achievements of complex information system theory with the creative interpretation of the problem of “existence” of philosophy. And his definition of information is “Information is a philosophical category indicating indirect being. It is the self-manifestation of the existing mode and status of matter (direct being)” [13]. One of the core values of this definition is to establish the dual existence of matter information of the whole world and all things in the world. It is in the sense of dual existence of material information that we can say that system evolution is not only a material behavior but also an information behavior, and material system can be interpreted from the perspective of information system. Both the characteristics and evolution of self-organizing systems have “information characteristics.” This is the basic connotation of “information characteristics” of the evolution of self-organizing systems. (Wu & Nan, 2019, p. 1)

To be fair, leading Western philosophers of information have also asserted that information is the ontic bottom of everything (Floridi). Others have made the same claim about computation (digital ontology, or it-from-bit theses. See (Dodig Crnkovic, 2011) for a recent example). However, there is a lot more ontic baggage in the above assertions. To be equally fair “dual existence of matter information” is almost as incoherent – conceptually – as one can get. Moreover "information characteristics" are - what, exactly? Matter being matter? Matter being matter indirectly? It's enough to make one list questions.

Could it be that the "indirect being" Nan and Wu refer to is the same exact thing as Deacon’s absential content: that the nature of information – or the information content of a system, signal, or structure – is given by what’s absent? It certainly could be, and probably is. Frankly, however, it’s difficult to tell (you won’t know for sure after reading their paper), which puts Wu in good (largely) a-priorist metaphysical grand-schematic company – with the likes of Heidegger, Husserl, and Kant (and Deacon, and Hofkirchner, (2009)). Using the principle of charity and keeping in mind the language gap, we’ll just assume that “Information is a philosophical category” and “dual existence of matter information” is just a mistake of translation (otherwise there’s more trouble afoot than can be dealt with here.) Presumably “…self-manifestation of the existing mode and status of matter (direct being)” might be a reference to what Deacon, using his modal-absential conception of information content, would call what is present, as opposed to what’s absent (where what’s absent is both causal and the source of external work that forms the basis of information).

To be more fair, Wu and Nan do intend to some extent to “read [metaphysics] off our best scientific theories” as Steven French puts his formula for scientific ‘viking raider’ metaphysics. However, commitments to a-priori conceptual analysis bring the effort largely undone. For example - a commitment is made to a kind of materialistic dualism – or perhaps I should call it (wondrously contradictorily) a monistic dualism.

Søren Brier and Zhou provide us with an attempt at a detailed analysis of what is going on with Wu’s grand-schematic ontology, culminating in a recommendation that what Wu is mostly missing is – err – Peircian semiotic insights i.e. Another grand schematic ontology that is even older than the Leninist dialectics of nature! :

The subjectively experienced reflection in the water is somehow objective, independent of consciousness. But the reflection of moon is not moon itself. Wu thinks that this part of the field of existence, namely objective unreality, is objective information. It is the information world, which has been overlooked by past philosophies. Brier (2008, 2013b) would call it signs, not information, as information is only one aspect of the Peircean sign function. (Liqian & Brier, 2015)

(Note that Brier’s commitment to Peircian semiotics seems at first blush to be an inversion of that pursued by Deacon. Brier makes information a necessary condition of signs, whereas Deacon makes a Peircian sign-interpretant a necessary condition of information.)

On the one hand Wu’s direction is perhaps unsurprising given all of the talk in the philosophy of information about absence, modal semantic content, and theories of semantic information. However, it’s not quite the kind of raiding of stores of a-priori concepts that Steven French had in mind when he suggested Viking conceptual raids on those stores by aspiring scientific metaphysicians to find what works with empirical findings (Bueno, 2019; French, 2014; Ross, 2016). The inherent contradictoriness of the concept of materialist dualism (or at least material duality) is enough to deliver that conclusion. It’s no explanation to say that “system evolution is not only a material behaviour but also an information behaviour” – not without a very explicit conception of information given up front. It’s just trivial (and no reason for odd dualist commitments) that “a material system can be interpreted from the perspective of information system.” It’s possible that what’s being sought is something like David Chalmers’ double aspect principle for phenomenal and physical information, but this is difficult or impossible to discern from the description, or lack thereof, provided by Wu.

The distinction between Shannon information and physical causation is well understood, but it is not broadly ratified and there are significant problems with it with respect to individual signal content, semantic information, and transmission of true messages (Dretske, 1983). The same goes for the relationship between informational structure and physical structure. In the case of Floridi’s informational structural realism (ISR), transcendental differentiae de re – the basis of information – are physical differentiae in re when embedded in physical ‘spaces’ or structures. Moreover, Floridi’s Kantian transcendental conception does not preclude the possibility that all differentiae de re (non-uniform binary relations) in all kinds of spaces somehow reduce to, or else strongly supervene upon, physical differentiae de re at some (lower) level of abstraction. This is especially true for the information in natural phenomena.

Just when one thinks one has a handle on the striking fecundity of conceptualisation and ideation streaming from Wu’s article, there’s this clanger to deal with:

Here, the order pattern emphasized by self-organization can only be established in the sense of stable maintenance or complex evolution of information activity pattern. Order is a framework for encoding information. The maintenance of ordered structure is relative to the stability of specific information coding architecture. In dynamic order, the aspect that marks the essential invariance of a system is neither the quality nor the energy of the system. The quality and energy of the system are always flowing and changing, and only the architecture of the corresponding order of specific information coding is the sign of the nature of the system, because only this architecture is stable. A complex evolution of an ordered structure is a complex reorganization or construction relative to a specific information coding framework (more orderly new structure generated), the degree of system self organization growth and the increase of the complexity, that is, transitional evolution from an ordered state to a more ordered state are based on the complex reorganization or construction of a specific information coding framework. (Wu & Nan, 2019, p. 4)

There’s so much process-ontological and systems-theoretic hand-flailing here that I only feel qualified to attempt to (almost) address it because in my thesis (and forthcoming book) I present a basic theory of natural encoding for natural dynamical processes systems. Let me say only this at this point, with respect to the above quote: it’s unclear what a code is, and it’s especially a puzzle how exactly “order is a framework for encoding information”. At minimum, this idea alone requires enormous explanation and justification. It gets neither (although there’s lots more terminologically dense conceptual flailing).

The debate about the nature of codes, and whether one should be realist about them, in the philosophy of biology alone is enough to make the boldest grand-schemer wary. Or, it should be: (Barbieri, 2013, 2018; Bergstrom & Rosvall, 2011; Godfrey-Smith, 2000; Sarkar, 2018; Shea, 2007). This is even more relevant to sections 4.3-5 of the paper wherein the concept of a codon is extrapolated from its meaning in genetics to be applied to general replicating features (Dawkinsian replicators are called into service too) of what Deacon would call autogens (and in fact the concept is used to describe exactly the same kinds of autopoietic non-organismic and organismic mechanisms.)

There’s a clearly important concept that Wu and Nan call ‘information mode selection’, but I am not sure there is a philosopher alive that could tell one what this is intended – exactly and accurately – to mean within the overall framework. It’s prima facie something to do with the specific systems-theoretic application (cybernetics versus chaos theory, for example) currently of interest. Why one needs a unifying conception of information that works on this basis is not made clear. Maybe it would be clearer if Wu actually said what information is beyond it being a massy-mode-ey “philosophical category” with monistic-dualist Leninist material features (I am of course being a little – err – 'flamboyant' here). [Maybe Aunty and Uncle Wang, or Uncle Hippie, know?]

[Okay – I’ll make the following observation also. It looks like a kindly but misguided (and somewhat evangelical) Deacon (or perhaps his acolyte Zhou) have been on a telecall with postgrad Nan to feed her the good oil. A fantastical miniature narrative perhaps, but, like I said – academic relationships.]

For all the common ground that Wu and Nan apparently (recently) share with Deacon (and his acolyte Zhou), it’s fair to say that they also have some large disagreements. Wu and Nan embrace a dualist outcome. Deacon (and Zhou) reject any such thing wholeheartedly:

[E]fforts to integrate physical theories able to account for spontaneous order with theories of mental causality end up positing a sort of methodological dualism. Simply asserting this necessary unity—that an observing subject must be a physical system with a self-referential character—avoids the implicit absurdity of denying absential phenomena, and yet it defines them out of existence. (T. Deacon, 2012, l. 279)

Zhou is right about many of the shortcomings of Wu’s work. However, Zhou’s own solution is a way out of the ontic frying pan into the explanatory and a-priori conceptual fire. Here is a reasonably lengthy excerpt from the same paper which outlines Zhou’s rejection of Wu’s approach as too objectivist and reductionist and indicates Zhou’s subjectivism about the nature of information. (Note the serial questions, which approach is common in Zhou’s philosophy, and commonly understood to be bad practice. Philosophy is certainly about asking questions, but the investigation thereof is by way of argument – not by way of just listing questions):

It is therefore evident to many researchers that an evolutionary theory of information, cognition, meaning, consciousness and communication put certain demands on the ontological presumptions of nature by science (Küppers, 1990). However, even if we believe in emergence, it is difficult to take as a point of departure a paradigm of nature based on an ontological materialism that sees nature and the emergence of conscious man as completely determined by absolute and universal natural laws. Actually a theory of emergence is not compatible with mechanical materialist determinism as it is based on a reversible time concept with no real irreversibility, and new levels of complexity as Prigogine managed to describe them in his non-equilibrium thermodynamics (Prigogine & Stengers, 1984; Prigogine, 1996). But it is, on the other hand, an intrinsic part of dialectical materialism (Creaven, 2007). As a replacement for the reductionist physicalist scientific understanding of nature in the West for instance, Fink (2006) suggests developing important points of McDowell’s (1998) book Mind, Value and Reality, into an transdisciplinary or absolute naturalism that does not see culture, mind and ethics to be outside nature. They are all natural phenomena and therefore inside nature. What else can they be, when modern science becomes more and based on a monistic evolutionary process philosophy and does not work with any form of absolute dualism? If we are not dualist then there can be no absolutely qualitatively different worlds inside and outside human mind and nature at large and thus we will have to enlarge our ontology to include values, consciousness, meaning and language. They can all be seen as aspects of nature’s great evolution. Why should the physically describable aspect of nature be more fundamental than the qualitatively first person experiences or the social intersubjective? Why should matter, energy and objective information be the only basic constituents of reality/existence? An alternative solution is that the standard physicalistic understanding of cosmogony and human evolution has to be revised on a more transdisciplinary basis. One attempt is that of Deacon (2012) that attempts to integrate thermodynamically based self-organization theory with information science and Peircean semiotics and a new view on the causality of the void leading to a new understanding of the evolution of mind. (Liqian & Brier, 2015)

Obvious extreme ‘liberal’ naturalism aside, Zhou’s work also includes careless and too-quick statements and strong assertions about both the nature of information and the history of the philosophy of information in the informational turn. Emergence, for example, constitutes perilous ontological ground indeed, and that’s before one even gets near emergent information. It may transpire that the whole question of emergence rests upon the truth or falsity of certain kinds of ontic anti-reductionism (Fodor, 2012). Maybe emergent information will explain emergence, but probably not before we all know what information is (or at least before physicists do). It may turn out that what some want to call emergent properties and behaviours are in fact objectively reducible but on a completely epistemically inaccessible basis. Lack of epistemic access does not entail irreducibility. It just entails - lack of epistemic access. This is for precisely the reasons that classical information theory avails us of: noise, signal loss, channel limitations, and signal equivocation (which incidentally provide a pretty good explanation for the necessary partialness of all internal and external representations).

Zhou tells us that “Norbert Wiener established cybernetics and united information theory and thermodynamics”. This is presumably an allusion to Wiener’s famous methodology. However, it brings with it some notorious baggage: the idea that Shannon entropy and Boltzmann, or else Gibb, entropy, are the same, or identical in some important physico-mathematical sense. This is at best an open question.

Zhou’s recent work on applying epistemic and explanatory complementarity in the philosophy of information is more thorough, but needlessly otiose, and based on synthesis of a narrow set of premises. It perpetuates some of the same errors as Deacon’s work: the wrong assumption that subjectivism about information is true (T. W. Deacon, 2013, pp. 392–393), and the thesis that Shannon entropy and Boltzmann entropy are not just related, but in important ways identical. The latter thesis has been debated for some time, and at best it is still and open question. In fact it is probably wrong (see previous article).

Complementarity is Zhou’s weapon of choice for solving the problem of unification in the face of pluralism (Zhou, 2020). Zhou’s attempt to use it to defend an interdisciplinarity commitment for the philosophy of information to support Deacon’s merging of dated and comparatively arcane Peircian semiotics with classical conceptions of entropy in information theory and thermodynamics (see previous article #). Zhou says that “the idea implied by complementarity does support transdisciplinary theories.” (Zhou, 2020, p. 294) However, the merging of Peircian triadic semiotics and its associated metaphysical commitments with classical Shannon and Boltzmann entropy really are not apt for application of complementarity.

Neither interdisciplinarity or complementarity save the phenomena, or the nature of information and a unificatory concept of it, here, nor could they. Complementarity was designed for reconciling concepts for interdisciplinarity in the hard sciences, and it’s not even assured of working in that setting. It’s not even purposed for justifying the combination of arcane (Peircian) semiotic, cosmological, and metaphysical theories with hard and applied science and mathematics. Trying to make it do so is not convincing, and does not deliver the Deaconesque style of grand-schematic information philosophy from its incoherencies, conceptual otiosity, unnecessary ontology inflation, and a-priorist whimsy. It possibly more successfully applicable (though not much more) to Wu and Nan’s parallel comparison of various research ‘creeds’ (systems theory, cybernetics, chaos theory, complexity theory, information) for the purposes of informational unification (keeping in mind that there are two kinds of unification in play – one about the nature of information, and one that uses information as a unifying element for theory unification. See the end of section 3.6 of Wu & Nan, 2019, p. 6). I think that creeds have something to do with modes. Probably.

The largely inherited (from Deacon) subjectivist-interpretationist (about the nature of information) perspective is probably the most egregious flaw in Zhou’s work. It is not just that thoroughgoing subjectivism – in Deacon’s case a broad naturalistic non-cognitivist interpretationism - about the nature of information does not respect pluralism in the right way. It’s that it ignores the intuitions and working assumptions of most working scientists – that information can and does exist objectively and without any need for interpretation, receipt, or functionally orientated or evolved consumption. Without, dare I proffer, any Peirce-worshipping semiotics, or even any signs for that matter. The lesson is – if you want to grand-unify in the face of pluralism about information, don’t use semiotics and necessary interpreters to do it.

[You can stamp your foot and demand that I respect Peirce’s authority if you like, but I am likely to respond by talking a lot about ontic structural realism and algorithmic information theory. Then I might really annoy you by talking about reduction.]

It’s true that in many scientific cases – especially in genetics and evolutionary biology – reception and definition of semantic or else semiotic informational content may be defined as involving a receiver, consumer, or interpreter on a telefunctional, teleosemantic, or even what Nicholas Shea has called an infotel semantic basis, because signals have evolved (as Deacon agrees) to serve an evolved function at the phenotype, or brain function, or whatever biological mechanism is the end product of the propagation of some signal, emission, or structure. However, it does not follow that this overrides pluralism about the nature of information, and entails that information is not also an objective commodity realised by uninterpreted and unconsumed causally-induced structures and patterns. Even more relevantly - subjectivism’s simply not coherent when applied to emissions and signals of and in natural phenomena. It doesn’t cohere well with what’s intuited and known, and with what classical information theory says about, the information of systems and phenomena such as celestial X-Ray sources, electrochemical signalling in brains, and DNA translation-transcription and protein synthesis in cells, among many others.

An easy way to see this lack of coherence, and failure to ‘save the phenomena’, is to realise that most scientists and folk would regard that a signal or an emission contains information about and from its source in the absence of any receiving agent. This intuition is sound, and does reconcile with both classical and algorithmic information theory. It’s also a far better respecter of ontic parsimony (Ockham’s Razor), which matters in scientific metaphysics as well as in a scientistic naturalistic metaphysics. A receiving agent is simply not a necessary condition for the obtaining of information.

[I suspect that one of the signals that the existence of say - celestial X-Ray sources - sends us as philosophers of information might be something like ‘sign-shmine’.]

Keep in mind that, importantly, information is not identical to information transmission. That would be conceptually, definitionally, explanatorily, and ontologically circular. Certainly an interpreting receiving agent is not a necessary condition for information. Thus claiming that a receiver-interpreter is a necessary condition for the information content of a signal – and of the source which generated it – is best said to be wrong. Unification in the face of pluralism about the nature of information cannot involve making all information dependent on an interpreter.

[I think that we philosophers of information are allowed to have more reductionism than that – just as a treat. (Physicists wont mind sharing some of theirs.)]

Deacon attempts to sidestep these interpretant-cramming problems by using a baroque schema involving what he calls teleodynamic systems and their simpler autopoietic counterparts – autogens (Deacon, 2008, p. 193, 2013, pp. 109–110, 223, 226–228, 311-313) (the hierarchy is more complex than this, and requires a large amount of verbiage to even describe, largely because Deacon has a grand-schematic penchant for copious technical neologisms which has troubled numerous reviewers (Fodor, 2012; Mcginn, 2012)). Nature is made up of simple and autopoietic systems of various levels of complexity. Those that arise by processes of biological evolution (and in some cases simpler evolved chemical systems) are attributed with a recursively self-reinforcing interpretant, and are called contragrade (Wu and Nan talk about the breaking of the “inertia” and stability of the structure of an existing system to perpetuate the orderly evolution of a system).

In the previous article I mentioned that this addition of a necessary per-system evolved auto-interpretant to the ontology is due to Deacon’s commitment to reconciling his metaphysics with the esoteric and dated metaphysics and accompanying semiotics of Charles Sanders Peirce. This is alone a dubious strategy (I submit that it is just wrong-headed), and not good scientific metaphysics. It seems to be the primary motivation for forcing subjectivist interpretationism as a necessary condition for the obtaining of information. Such an innate interpretant is not necessary at all, because we do not need to impose semiotics upon nature for natural information. Crow-barring an interpretant in to very simple dynamical systems and their dynamical structures, and along with it a whole Peircian triadic semiotic framework, bends Deacon’s ontology and theory into all kinds of uncomfortable explanatory and metaphysical pretzels.

These problems alone do not help Zhou’s piggybacking approach (piggybacking other philosophers is no mortal sin in and of itself, and a practice that is common in philosophy), but Zhou has many other difficulties due to trying to follow Deacon in making huge grand-schematic proclamations. For example, his predilection for Collier’s conception of information as reducible to symmetry breaking in nature is largely irreconcilable with the abovementioned subjectivist and Peircian approach by Deacon (Collier, 1996). Collier essentially says that natural systems are informational to the extent that they are asymmetric, but there’s little need for any recursively re-enforcing Peircian interpretant. None, in fact.

The current offerings from grand-schematic unifying philosophies of information from Chinese scholars remind me of this salutary warning from Jerry Fodor, delivered as part of his review of Deacon’s book:

In general, epistemological and metaphysical issues tend to be matters of some delicacy. It’s almost always a mistake to call in the heavy artillery right at the start.

This is not a warning that grand-schematic philosophers of information – especially those in The Middle Kingdom – seem interested in. I think this is to their obvious detriment. Since Zhou is following Deacon as acolyte, and Wu and Nan are following him as dualist and anti-subjectivist competitors, I thought it fair to let Fodor have the last word:

Consciousness, intensionality and teleology are, after all, things that happen routinely at the macrolevel (and, quite possibly, nowhere else). It’s hard to believe that saying what they are and how they happen requires lots of heavy-duty stuff about thermodynamic equilibria, neural attractors, morphodynamic activity, self-organising systems or, heaven help us, the quantum mechanical collapse of probability waves; nothing Deacon says explains why it should.

As it is for these explanatory targets, so it is for information. I recommend, and have always recommended, a more reductionism-friendly (or at least tolerant) physicalist approach to the nature of information in nature that defers to ontic structural realist premises which allow for dynamical systems and processes and their information content. That’s still my disposition. Process ontology is still allowed if one must have it, although, honestly, it looks like science is telling us that processes reduce to perturbations in the structure of the quantum field. Are perturbations processes? It's probably important to know if we want to find all of the classical information sources - which are stochastic physical processes. We can keep systems and mechanisms, but at a higher level of abstraction and with causal structure as the reductive base and ontic determiner. Grand metaphysical schemas are not what scientists usually do, which is why they end up attempting grand unifications. Naturalistic philosophers should take note. Scientists do attempt grand unifications that usually have reduction of some kind as a premise. Moreover, they understand what Shannon understood, and Fodor intuited – that diligence is required because:

“Seldom do more than a few of nature's secrets give way at one time.” (Shannon, 1956).


  1. Adriaans, P. (2010). A Critical Analysis of Floridi’s Theory of Semantic Information. Knowledge, Technology & Policy.

  2. Aleskovskii, V. B. (2002). Information as a factor of self-organization and organization of matter. Russian Journal of General Chemistry.

  3. Arenhart, J. R. B., & Bueno, O. (2014). Structural realism and the nature of structure. European Journal for Philosophy of Science.

  4. Barbieri, M. (2013). Organic Semiosis and Peircean Semiosis. Biosemiotics.

  5. Barbieri, M. (2018). What is code biology? In BioSystems.

  6. Berenstain, N., & Ladyman, J. (2012). Ontic Structural Realism and Modality.

  7. Bergstrom, C. T., & Rosvall, M. (2011). The transmission sense of information. Biology & Philosophy, 26(2), 159–176.

  8. Bueno, O. (2019). Structural realism, mathematics, and ontology. Studies in History and Philosophy of Science Part A.

  9. Collier, J. (1996). Information originates in symmetry breaking. Symmetry: Science and Culure.

  10. Deacon, T. (2012). Incomplete nature: how mind emerged from matter. Choice Reviews Online.

  11. Deacon, T. W. (2008). Shannon - Boltzmann - Darwin: Redefining information (Part II). Cognitive Semiotics.

  12. Deacon, T. W. (2013). Incomplete nature : how mind emerged from matter (1st ed.). W.W. Norton & Co.

  13. Dodig Crnkovic, G. (2011). Dynamics of Information as Natural Computation. Information.

  14. Dretske, F. (1983). Précis of Knowledge and the Flow of Information. Behavioral and Brain Sciences, 6(1), 55–63.

  15. Esfeld, M. (2017). A proposal for a minimalist ontology.

  16. Floridi, L. (2008). A defence of informational structural realism. Synthese.

  17. Fodor, J. (2012, May 24). What are trees about? London Review of Books, 1–5.

  18. French, S. (2014). The Structure of the World. In The Structure of the World.

  19. French, S., & Ladyman, J. (2010). In Defence of Ontic Structural Realism. In Scientific Structuralism.

  20. Godfrey-Smith, P. (2000). On the theoretical role of “genetic coding.” Philosophy of Science.

  21. Hofkirchner, W. (2009). How to achieve a unified theory of information. TripleC: Communication, Capitalism & Critique. Open Access Journal for a Global Sustainable Information Society.

  22. Liqian, Z., & Brier, S. (2015). The Metaphysics of Chinese Information Philosophy. Cybernetics & Human Knowing.

  23. Long, B. (2018). ISR is Still a Digital Ontology. Erkenntnis.

  24. Mcginn, C. (2012). Can Anything Emerge from Nothing? The New York Review of Books.

  25. Psillos, S. (2006). The structure, the whole structure, and nothing but the structure? Philosophy of Science.

  26. Ross, D. (2016). Vikings or Normans? The Radicalism of Naturalized Metaphysics. In Metaphysica.

  27. Sarkar, S. (2018). Decoding “Coding.” In Molecular Models of Life.

  28. Shannon, C. (1956). The bandwagon ({Edtl}.). IRE Transactions on Information Theory, 2(1), 3.

  29. Shea, N. (2007). Representation in the genome and in other inheritance systems. Biology & Philosophy, 22(3), 313.

  30. Wang, W. (2008). A critical analysis of structural realism. Frontiers of Philosophy in China.

  31. Wu, K., & Nan, Q. (2019). Information Characteristics, Processes, and Mechanisms of Self-Organization Evolution. Complexity.

  32. Zhou, L. (2020). Complementarity in information studies. Synthese.

Quine in Tibet?