Comments on ENSO Seminar “Radical Embodiment and Real Cognition”

Over at 4e Cognition Group Anthony Chemero has given a talk (YouTube link) about a couple of interesting new directions that he and his students are working on for their dissertations and a paper. The main impetus is to explain “higher order cognition” through a rECS-able perspective.

The first turn is through Gui Sanches de Oliveira’s Artifactualism approach to models, essentially giving a thorough and solid argument for that scientific models are foremost tools, not accurate representations of the world. If it works, we use the model to predict, explain, plan, experiment, etc. It reminds me of the futile path that scientists often are found on: Focusing on finding The Truth, or finding objectivity. But the world seems to me to contain none, but even if it does, it doesn’t matter, at least not nearly as much as if the proposed model can be used in any applied setting. It reminds me of Nancy Cartwright’s arguments about truth and explanation, how far away those two concepts are from each other -and opting for truth takes us further away from a functioning tool. This is a really important step. Artifactualism rightfully criticizes the assumption that thoughts are for representing the world accurately, and replaces it with that cognition is for toolmaking. “Explicit, occurrent thoughts are tools, instruments, or artifacts that some agents create and use. Of course, models can meet formal definitions of representations, but that is not what they are for…”.

The second turn is through Vicente Raja Galian’s attempt at defining brain activity through resonance and oscillators. In his case, TALONs as resonant networks of neurons that resonate to certain ecological information and not others, that can continue to oscillate in the absence of the initial resonate -and that can be set in oscillatory motion at a later point in time (again without the initial resonate, through Ed Large’s work). The brain here, is driven by everything else (not the opposite way around). Oscillators, and non-linear oscillators, can act as filters and produce patterns not in the original driver.

Then, we take a turn into what Chemero refers to as slave/master systems, and while those words seem very culturally loaded, they make the point that slave systems wander (drift) in absence of a master system. E.g. circadian rythms stay in tune when we are regularly exposed to sunlight, but when deprived, our rhythms start to drift. An idea connected to that when we do try and use TALONs to think about things, or the past, but because it is not what they (and as a whole, the brain) is for, we just don’t seem to be very good at it. Marek McGann adds “‘Memories’ are constructed on the fly, and confabulation is rife, because it is not retrieval of things, but it is temporary toolmaking”.

Ultimately these initial steps in making more concrete the idea of ‘resonance’, seems very promising. An interesting aspect of resonance, is that it exists on all scales, it doesn’t matter if we look at the behavioral or neural scale, which makes them analyzable by methods like fractals. It makes it an empirically testable theory. Also, with resonant networks, they no longer have to contain content -Anthony Chemero suggests tool-making which will have to be defined further for me to understand if representational content hasn’t just been replaced by Gibsonian tool content. And don’t get me wrong, that would be a wonderful first step in better characterizing what humans do, but I am also currently on a quest for a non-content description of neural activity -and resonance seems to fit that description.

Unofficial lecture on representations, intro to rECS and Master Thesis

So I’m writing my thesis on abandoning representations and replacing it with ecological psychology, and this is bits and pieces of what I’m writing. To fit one lecture I obviously had to leave out a whole lot of information. Even information that would change some of the subject matter. The idea I had was to introduce, not even all of, the basic stuff I have in my thesis and was hoping to get some critique and comments on it.

Link to video;

Most sources used in the video;

Blogs and blogposts
Scandinavia And The World (illustrations);
Eric Charles blog post;
Wilson and Golonka’s blog;

James Gibson – The ecological approach to visual perception
Anthony Chemero – Radical embodied cognitive science
Pfeifer and Bongard – How the body shapes the way we think
Gerd Gigerenzer – Rationality for mortals
Bem and de Jong – Theoretical issues in psychology

Tim van Gelder – What might cognition be if not computation
Fodor and Pylyshyn – Connectionism and cognitive architecture: A critical analysis

Virtual affordances. Electronic Sports (and Computer Resistence).

I’ve grown up with computers since birth, in fact, one of the first generations to do so. The virtual world needs to be accounted for, but I accept the non-affordance of pictures, depictions, movies and thus screens overall (but see for very recent, in-depth, information) I believe it is enough to denote this with the word virtual. This is something I go to some length with in my thesis and the beneath is the preliminary version of that section. While there is some revision still to take place, the main content is there and should provide enough clarity as to what I mean by it.

The world of electronic sports (henceforth; e-sports) is a largely unexplored area even within traditional cognitive psychology. In rECS it is discounted, essentially, because it is performed on a screen and as such does not provide affordances per say. In agreement with this, you still cannot just ignore this massive field. It is not only entertainment, it is for some a way of life and it is for others their monthly income -both as creators as well as players. In an attempt to refrain from legitimising the field further, it stands for itself in the amount of hours played, the number of games produced, the amount of profit for gaming-companies and the prize-pools for e-sports players. One aspect however, that is unstated in the relevant literature, is that unbeknownst to producers and programmers of games, their absolute central aspects, follow exactly that of ecological psychology and rECS. Gibson (1986) made the same analogy, however with greater depth, for the fields of architecture and design.

A programmer creates the environment in which a player is to exist and, hopefully, immerse herself. The virtual environment is created in respect to contain virtual affordances for the player, or for the player to explore and act within. The evolution of computers, as well as the games played on these computers, have increasingly dealt with the fact that players expect more and more virtual affordances to be available to them. There is an expectation to be able to do more things, to increase the complexity of the virtual environment, virtual objects and other virtual agents. When expansions are released for already popular games, they account for this fact by not only adding new items, for example in MMORPGS (massively multiplayer online role-playing games), but also by creating new virtual affordances to players through new game modes (changing virtual affordances of the already known game), allowing completely new virtual behaviours and thus making the virtual environment increasingly complex. For games that insist on reflecting reality, the expectation is that virtual affordances should more and more closely resemble the environment. This is thus an essential area to account for when it comes to rECS and psychology in general. It is necessary however to introduce the term virtual affordances, because as stated, pictures, depictions and even movies do not present affordances (Gibson, 1986 and Wilson & Golonka, 2013). Nevertheless, computer gaming industry works with manipulation of virtual affordances, and thus, virtual affordances are defined as invariants programmed in environment, objects and agents, allowing, limiting or disallowing virtual behaviours, interactions and coupled systems between those environments, objects and agents.

The game of choice for exemplification, is League of Legends (launched 2009, by Riot Games, formed in 2006). It is played by 32 million unique players every month, 12 million of which play daily, racking up 1+ billion hours of play each month making it the most played computer game in the world (Riot Games, 2012). They have created a virtual environment in which there is an economic system; killing AI-agents and opponents grants money, from which you may buy items to further enhance your characters basic, level-dependent, properties. The virtual environment affords movement in two dimensions but also limits movement by walls and shrubbery. Each character, 110+ to choose from, is afforded five specific abilities (one passive, meaning it is not “usable” by pressing a button and four active abilities assigned to one key each) plus the choice of two out of thirteen that are common to all players. Some abilities modify movement capability of oneself, of other agents, amount of damage given, amount of damage taken and/or regeneration of vitals (health, mana or for a few characters, a specific other vital coupled to its offensive and/or defensive abilities). Two teams with five players on each team thus comprises (10*5*2) 100 agent-specific virtual affordances, coupled with the dynamic variety in which the virtual environment lends itself to each specific character. Needless to say, perceiving one’s own and other characters’ virtual affordances, in which sequence they are used, and in which situation, the dynamic relationship all these variables have in conjunction with where one is situated in the environment and your, and their, vitals, is what counts as skill in this game. It is a visual perception heavy game but auditory perception enables you to gain information on parts of the environment not currently in your virtual visual field but that may have an impact on your virtual behaviour. The mentioned variables are far from an exhaustive list; there are quite many more virtual affordances to be described, but these should be enough for even the most computer-illiterate to understand that it is far from a simple virtual environment to navigate through successfully. Thus, this complexity gives rise to a vast range of behaviours and emotions, one of the most extreme of which is called “rage-quitting”. It is when you are sufficiently angry, regardless of why, that you exit the game before completion and leave your team severely underpowered against the opponents. Similarly, it is what can be seen in real life interviews when interviewees physically leave the interview prematurely.

Computer-gaming, although not adhering to the strict definitions of rECS, needs to be accounted for and it is suggested that it is sufficient to discriminate between real life and gaming by the verbal notation virtual. When experimentally reporting on computer games or screen-dependent research, it is of great importance to include an exhaustive list of variables and virtual affordances in the previously mentioned task analysis. This leads on to the study at hand, where an attempt is made to follow this task analysis for rECS experimentation, in order to show its practical application; to try and create headway for computer-screen experimentation by refuting the unwillingness within the embodied perspective towards it; discriminate between predictive and prospective strategies in problem-solving to discriminate between computational and ecological strategies; and illuminate how lucrative future research can be on the basis of both the process under observation and, more generally, to produce knowledge about it through rECS.

Communicationally necessary separation of objective and subjective perspectives (in rECS) (1/3)

I began writing the situated relationships between the concepts (mentioned in my previous post) and realised something terribly important. Even in the simplified taxonomy, I haven’t separated out subjective from objective, and I found out just how important that is when writing about the specific relationships. They exist in different realms (akin to the ontological and epistemological issues I have been writing about), also, communicating subjective relationships will depend on the specific organism and its umwelt (Louise Barrett). I have, for now, had human activity in mind, in an effort to keep it simple. This will guide the way I henceforth communicate about relationships in rECS where necessary to specify, unless someone has a good reason not to…

Objectively, here, refers to a mind-independent, theoretical perspective. I am not concerned here on how we come in contact, how we experience the world, but rather on the relationships between the concepts in how they affect each other, separated from how they are experienced (or might be experienced). It is not to do with separating ontology from epistemology, but there are surface similarities. For example, talking about Realisation and Actualisation, in an objective perspective you cannot have Actualisation without Realisation (I have written otherwise in other places, and should be revised on the basis of not separating objective and subjective perspectives clearly). This is so because Realisation is defined as perception of affordances, and, you cannot interact, act on, Actualise, affordances without perceiving them. The same goes for Limitations, which may be present and affect Actualisation, but not necessarily be experienced.

But. In a subjective perspective, here defined as experiential, i.e. how we experience the world. We can Actualise affordances without “paying attention” or consciously or deliberately perceive, we just act. An example can be very quick decisions, we need not experience the Realisation of the acted on affordances. Again, in a theoretical sense, an objective perspective, it is clear that we have to have Realisation (perception of) on some level, whatever level that is, for us to be able to Actualise the intersituational-affordances-relationships. Reflexive behaviour could exemplify this, since they are usually experientially Realised after one begins Actualising, after the affordances have been Actualised or not at all.

Thus, it is important to create two separate taxonomies for experiential, subjective, relationships (which will become mostly an empirical endeavour to sort out experimentally) and another for theoretical, objective, relationships. The theoretical perspective will necessarily incorporate more aspects, more relationships and be truer to dynamic systems theory than the experiential perspective. This is explained by the examples above and by that what we experience is dependent on our senses, which obviously are “limited” (put in quotation marks because I do not wish to support the view that we ought to be ideal agents, should be measured on the basis of ideals or are heading that way through evolution, since this imposes a frame-of-reference error. We are humans, and have developed under the pressures of our environment, and this is what we are, nothing more and nothing less).

If I find the time to explicate those taxonomies is another question…

Simplified taxonomy of modified rECS (5/5)

Well well, this is how far I’ve come in trying to visualise the whole tree of concepts in the modified version of rECS (Chemero), with additions from Golonka & Wilson and myself.

Starting out in the bottom right, with energy array and physical properties, it is worth mentioning that an energy array also could be said to be physical properties since we are talking about for example visually, light particles/waves. They are separated due to their function.

Energy array + Physical properties give rise to Structure.

Structure, non-perceived, is not information.

Structure + Perception give rise to Information.

Information give rise to Affordances of the object/agent and the Limitations.

Limitations + Affordances can be Realised and/or Actualised.

Affordances can be Realised and/or Actualised (without the need of perceiving Limitations).

Affordances can be Realised which can give rise to Actualisation.

Affordances can be Actualised giving rise to Realisation.

These are not static one-way relationships, change in one, changes the others down to Perception. Practically, there should be arrows from Affordances, Realisation, Actualisation, Limitations, Information and Perception, to each other.. My MS Paint skills need a bit of retouching for that to happen. I am on my way of separating out all the concepts one by one and link them to their implicated and or necessary concepts. This is meant as a simple overview.

Watch this space as I will try and post a new blog post each day (roughly) for each concept.