Rationalizing About Rationality: Ecological Rationality


The Thinker by Auguste Rodin

In this paper I will prove nothing less and nothing more than the following statement: rationalizing about rationality only stops when one gets tired or bored of it or when something in our environment joggs us out of the recursive chore of rationalizing about, rationalizations about, rationalizations about, ….

Part I

Now, before I go on, I’d like to provide some preview as well as context on why I’m about to spin spools and spools of yarn. This paper has been prompted by the write ups of Gerd Gigerenzer [1] and David Matheson [2] concerning philosophical inquiries into the notion of human cognition. So I’ll review their respective papers in Contemporary Debates in Cognitive Science and then present my rationalizations about rationality which borrows heavily from Gigerenzer’s overall work regarding the subject. My opinions are also informed by my professional experience in cognitive systems design, human performance measurements and human factors studies as a cognitive engineer. I will eventually present the philosophies of cognitive systems and frameworks of cognitive engineering I had mashed together for my sake which have enabled me to work as a cognitive engineer.

In my view, Gigerenzer actually presents five visions of rationality even though the title in his paper mentions four. So let’s count from the zeroth to the fourth vision of rationality presented by him.

The zeroth vision in Gigerenzer’s paper is that of 18th century scientist Pierre-Simon Laplace who came up with a thought experiment regarding an all knowing (omniscient) being that has unlimited computational power. This thought experiment is commonly known as Laplace’s Demon. The demon’s universe would obviously be deterministic and quite certainly the demon would never be surprised pleasantly or otherwise by impending future. The demon was supposed to be the ideal rational being which could deduce everything about its future based on its present while having perfect and thorough memory of its past. This type of theorizing was actually popular in those days and Gigerenzer points out that scholars attempted to compare human thinking to that ideal or “unbounded rationality” or “full rationality” and they even attempted to portray human beings as a model of that ideal.

Years passed by and people came to realize that we aren’t truly comparable to Laplace’s Demon and by the era of Daniel Bernoulli, views about human rationality were more in line with our experiential life which, is not deterministic, to say the least. This vision of rationality postulated that a rational human being’s rationality was exhibited in seeking out “optimality” of expected values and retained the notion of unbounded rationality. Of course, it seems to me that no one in those days asked how any sort of expectation come to arise in the first place. Gigerenzer considers optimization as a mental activity that seeks to find the best strategy to produce an outcome such that even if there are equally good strategies there aren’t any better ones. He also points out that scholars who subscribed to this vision assumed that persons or (in the view of economists) corporations had perfect knowledge of all the relevant options and potential consequences in order to formulate as well as further their desires.

By the mid 1900s, views about unbounded rationality became stale and faded from scholarly discourse because there were no reports of any human having thorough knowledge of all possible eventualities. The next vision about rationality which dropped the idea of omniscient determinism all together was concerned with optimization given the constraints of limited mental and/or environmental resources. This assumption of limited resources and performance capacities seemed more realistic. The notion of choice and how a rational human made a rational choice was an important concern for theorists of that era and it still is. Under this perspective, theories explored human behavior concerned with adaptation to natural constraints and uncertainty. The adaptation was supposed to optimize a hypothesized cost function and in some sense the behavioral adaptation was in itself optimal. According to me cost function, utility function and cost-benefit tradeoff allude to the same concept. Under constraints, the rationalizer had to introspect or prospect for information in order to make an informed decision and solve a problem they were experiencing. The challenge that those theorists never surmounted was in explaining how a cost function was built up in the first place, how exactly would it be resolved i.e. when would a rationalizer stop rationalizing about a decision and how the cost of evaluating a cost function was taken care of. This form of analysis about rationality quite literally lead to mental paralysis and in some ways the older notions of unbounded rationality seemed more appealing because they had simpler albeit unrealistic assumptions. One really important point that Gigerenzer makes at this juncture is that many scholars still confuse this vision of rationality concerned about optimization under constraints with “bounded rationality”.

So, here I’d like to mention that the term bounded rationality was (perhaps) coined and was indeed promoted by Herbert Simon who intended that the word “bounded” be used to explain the relation a human has with the environment while making decisions. Due to these relations the human settles for a satisfactory outcome of their decision even if they realize that the solution or outcome was not optimal for the given environment. I’d say that Simon did do the scholastic world a favor by stating that an interaction between a human’s capacities and their environment gives rise to outcomes. An anecdote about how Simon saw human capacity and the environment as two blades of a scissor that needed to come together in order to resolve a decision was reported by Gigerenzer and Reinhard Selten in their book – Bounded Rationality: The Adaptive Toolbox [3]. Simon pointed out the need to give credence to a human’s environment while investigating any mental aspect of that human for the environment possesses properties that allow for simplification of the decision maker’s choice mechanisms. Simon’s work also instructs us that we must admit to a “satisficer’s” subjectivity that decides what is or isn’t satisfactory.

This digression into Simon’s work was important for the next vision in Gigerenzer’s paper is related to the work done by his compatriots – Amos Tvrskey and Daniel Kahneman who built upon Simon’s work concerning heuristics i.e. simple choice mechanisms shaped by the environment. But unlike Gigerenzer and his colleagues who were doing the same during the 1990s, Tvrskey and Kahneman ended up winning a Nobel Prize in economics for their contributions to academia. I think the Nobel Prize was for the disruptive effect their work had in the schools of economics and the way in which they managed to dislodge the 18th century thinking concerning rational-economic-decision-making within those circles. Gigerenzer, of course, has not been satisfied with their work and points out that their discovery of how we make “systematic errors” or harbor “cognitive illusions” due to our biases is with respect to presumed norms of first order logic or theories of probability or (yet, again) optimization rules. Gigerenzer states that being able to perform first order logic or to conform to a given norm is neither necessary nor sufficient for being rational, specially if those norms do not have any ecological validity. Gigerenzer has been notably upset with how Kahneman and Tvrskey have interpreted Simon’s views and he states quite emphatically that other scholars have misunderstood the debate Gigerenzer has had with Kahneman and Tvrskey. The debate has not been about what is human irrationality or whether admitting to human biases requires us to admit to the notion that we aren’t rational in any sense of the word or that we are only rational to “some” extent. The debate has been about discovering that particular sense of the word called rationality which, allows us to scientifically explain, to ourselves, our thought processes concerning decision making.

Gigerenzer presents the fourth and final vision concerning rationality by steering the reader to his interpretation of Simon’s views and his own work considering bounded rationality where the word bound refers to the past tense of the word bind. In binded rationality the binding of human rationality is with respect to an environment and rationality exhibits itself within an observed ecology rather than within a conceptual analysis of human behavior with respect to a context-free norm. The anti-prescriptive English grammar in the previous sentence was necessary to explain how the word bound in this context is neither related to boundary i.e. a limit of something nor to bridle i.e. curbing something. And in my view this means that one must experience rationality within a relevant ecological context to know what it is rather than merely conceptualize about what it ought to be.

To better explain his notions about bounded rationality Gigerenzer first rebrands it as ecological rationality, possibly, to avoid the confusion that people have with Simon’s and Kahneman’s work. Then, he elicits the terms embodiment and situatedness. Next, he explains why context is important and how norms can be context-free and finally concentrates on examples of how heuristics have internal and external validities.

I urge the reader to go through the examples in Gigerenzer’s paper concerning heuristics and how those choice mechanisms are shaped by an environment. I believe, it would be considerably profitable to the reader to do so, especially, to understand counterintuitive phenomena such as: when one reason is better than many, when less is more and when ignorance pays. Without visiting those examples here, I’d like to present my interpretation of content-blind norms and then say a bit more about situatedness and embodiment.

Gigerenzer does pose the question, again and again in all of his works, “how does a norm come to exist in the first place?” He notices that norms or criteria can be arbitrary and applied to a decision without reference to the full context concerning the decision. He says that these norms are content-blind.

I believe that Gigerenzer means to say, “context-free” whenever he says, “content-blind”. Gigerenzer has discovered, for himself, that humans have an uncanny ability to regard a hypothesis as a law and evaluate subsequent observations based on that so called law. I believe that Gigerenzer has not fully reconciled some of the corollaries of Simion’s theories and his own views on ecologically embedded rationality; A norm that a human uses is never solely shaped by that human. Norms, I say, can never be context-free. Norms are idiomatic i.e. they have a relation to a speech community’s conceptualizations of the universe and to that community’s cultural evolution. Norms are opensourced and crowdsourced conventions formulated and accepted by a speech community to accommodate a very large variety of situations, such a large variety that the accepted norms may seem unfalsifiable or context-free. Also, one is never supposed to be able to achieve a norm in the first place because norms do not set a destination they set a direction. As such one never reaches “The East” one merely heads East while relying on their own, subjective notion of East. In a more literal sense of measuring East as a direction, there need not be, I’d even say there simply isn’t, an absolutely veridical and objective notion of East because one could always measure directions with respect to an arbitrary origin or datum and in n-dimensional vector space. The geographic notion of North, East, West and South as cardinal directions is an accepted convention which has certain limitations. It does not accommodate those points in space defined by an azimuthal as well as latitudinal coordinate while compensating for Earth’s elliptical shape. Even with its shortcomings one wouldn’t call that convention irrational because of its utility within certain navigational contexts for certain communities. As such, norms set the mood rather than a specific idea and while under the influence of such mood one is to explore and implement various ideas.

In the next few paragraphs I want to elaborate on the notion of ecologically embedded rationality to explore a means of generating norms. I don’t intend to define situatedness nor embodiment while doing so, I just want to muse about these concepts for a bit before commenting on David Matheson’s contribution to Contemporary Debates in Cognitive Science.

The situation that one is faced with is laden with information and is host to a multitude of free agents. It influences our ability to consciously or unconsciously cognize about future situations along with some hindsight. The situation does influence what we remember i.e. what is being accessed as hindsight. Our ability to observe ourselves and our universe is, in the philosophical sense, interval censored. Only an interval space or rather, spacetime, is a part of our situated cognition. Interval censored is a concept in statistical analysis which explains that our ability to observe an outcome may only inform us about the value of a measurand that could be at least a certain level and at most a certain level. For example, one might be able to think about their oldest memory but it certainly wouldn’t include what it felt like to be a zygote and undergo cell division. This makes our lives left censored. We also don’t exactly know what tomorrow would feel like even though we can make conjectures about it. This makes our lives right censored and as a culmination of being simultaneously left and right censored, our lives are interval censored… in the philosophical sense.

Whatever may be the mechanisms by which we can observe our semantic space and the natural universe within an interval of spacetime, those mechanisms allow us to flow into subsequent situations with or without an adequate understanding of the situation. But! A situation is certainly not dictated by our will alone as it is shared by countless free agents and entropy. So these mechanisms must include those agencies which are classified as “not self”.

One could certainly attempt to speculate about that which is beyond our censored (veiled) senses and sensibilities. However, I have always failed to understand what exactly do people mean when they say that they can introspect in order to achieve a realization of something to the fullest extent without experiencing that thing in its appropriate context via embodiment. For instance, I don’t see them manifest the consequences of being run over by a bus when I ask them to introspect about the situation. This indicates to me that they haven’t achieved full fidelity with what they were trying to speculate about. There are certainly quite a few things that we can only, at best, speculate about and the one who can tell the temperature of a flame without sticking his or her tongue into it might be particularly wise. But, perhaps we should simply throw some of these armchair-philosophers under a bus to let them know the difference between speculative and operative experiences.

Keeping the serious jokes aside, given our current knowledge about neuropsychology, it is possible that physicalists do have a significant point in saying that mental phenomena are explainable by physical phenomena using reductionism. However, it is also possible that these physicalists may have a certain physicality and/or have been through certain situations that permanently restrict them from understanding which portions of the mental are not reducible to the physical. This would be ironic, if true. Or alternatively, non-physicalists have a certain physicality and/or have been through a certain situation that permanently restricts them from understanding what physicalists have to say. This would merely be sad, if true.

So using Gigerenzer’s view on rationality or rather, my interpretations of it, I would say that Kahneman et al. are approaching the subject of rationality based on their semantic space which has idiomatically evolved within some cultural and environmental situations. Their semantic space has also been dependent on their physicality in order to achieve a certain value system through embodied cognition. This value system actually prescribes first order logic or bayesian statistics as criteria in evaluating a human’s decisions. So, all those who, like Kahneman, have tested human behavior with respect to first order logic or bayesian statistics are not terribly wrong in doing so. Of course, they are not terribly right either as their experiments do not have good external validity because they performed their tests only on adult humans and mostly in situations involving a business or managerial scenario. They wouldn’t possibly be able to comment on the rationality that babies and young children possess using their experimental designs which had helped them win the Nobel Prize in economics. They would need different experimental designs, ones with better representative design as well as stronger internal, external and ecological validities [see, 4]. And if one were to say that babies are inchoate and that rationality only manifests itself after a certain age in the form of critical thinking, then the question of why it is so has been completely missed by Kahneman and even Gigerenzer as well as their colleagues. So what did Gigerenzer manage to map out?

Taking the perspective elaborated by Gigerenzer, one can say that indeed, while rationalizing about anything, including rationality, there can be umpteen rationales. Also, there can always be a rationale for any human endeavor regardless of whether the person engaging in that endeavor is aware of that rationale. A separate observer who is observing such a person may also never become aware of such rationale. In that case, can a human ever be objective? This is discussed towards the end of this paper in an implicit way after commenting on David Matheson’s paper.

Part II

David Matheson also attempts to popularize what Gigerenzer has been ginning. However, the distinct flavor that he adds to the discussion is his acute focus on the information domain. He stresses on the fact that there needs to be a concordance between: the information that is available within an environment, the way it is retrieved by an observer for processing and how the observer’s algorithms influence what gets processed. Understanding the nature of the algorithms would tell us the types of rationales the observer can implement using the particular data types their algorithms were designed for, or rather, evolved for. Matheson quotes an analogy provided by Gigerenzer to explain this idea which, I’ll recount here:

"The algorithms of our pocket calculators are tuned to Arabic numbers as data inputs and would fail badly if we entered binary numbers. Similarly, the arithmetic algorithms acquired by humans are designed for particular representations.... Contemplate for a moment, long division in Roman numerals."

In case of humans and other organisms, we don’t really know what all those algorithms are and we can only see some sort of an output for certain inputs using some kind of a test scenario. This behaviorist paradigm of observing outputs that might be in response to inputs has become less fashionable specially due to the plethora of neuroscience research work like that of Alan Hodgkin and Andrew Huxley, providing highly corroborated evidence of what actually goes on within the organism which previously was treated as a mere locked-up-blackbox.

Which philosophy must one subscribe to concerning what is or isn’t {cognition, rationality, observation, understanding, beliefs, opinions, considerations, knowledge, emotions, sentiments, musings, thoughts, ideas, etc.}? The ones I subscribe to, for my sake, state that there are only beliefs concerning veracity and commensurability of beliefs. This belief is likely to be absent within the repertoire of those who can’t observe that veil which censors the rest of the universe while enveloping us within a situational paradigm. Everything and everyone beyond that veil or perceptual bubble, might as well be or not be {Schrodinger’s cat}. I believe that this sentiment was much better articulated in the canon of works produced by Paul Feyerabend, Michael Polanyi and Thomas Kuhn. Feyerabend and Kuhn quite naturally had slightly differing views on commensurability and Kuhn has revised his own views on the subject time and again but this is natural of us humans and our conditions.

Some people I’ve come across in my lifetime seem to reconcile with our natural functionalities which can max out, by not doubting anything that could be beyond the veil and by acting on resigned faith while others reconcile by doubting everything beyond the veil by default. I find the previous category of people “cognitively impenetrable”. But if you, the reader, ever come across a skeptic (a person who doubts things by default) and if they say things like, “skepticism is self-authenticating”, ask them to apply the concept of skepticism to that statement.

There are of course those who dynamically modulate their behavior contingent on their situation and intuition, doubting things only when there are sufficient cues or information that elicit doubt. These people are more pragmatic. To analyze this type of pragmatism is where Gigerenzer’s perspective along with systems perspective comes in handy. We can now articulate that particular dynamic modulation as the relation a cognitive system has with its inventory i.e. its embodiment as well as its environment, while being a constituent of the environment’s inventory. The circularity in the previous sentence is actually of paramount importance for this perspective allows us to observe and work upon the rationality of any cognitive system including the largest conceivable one – the universe.

What I term as inventory or repertoire is my interpretation of Gigerenzer’s “adaptive toolbox”. However, it might be inappropriate to attribute Gigerenzer with the perspective he has been popularising for it is possible that the credit, at least in part, lays on Egon Brunswik whom Gigerenzer et al. quote in their book – Simple Heuristics That Make Us Smart [5] as having said,

"Psychology has forgotten that it is a science of organism-environment relationships and has become a science of the organism.... This... is somewhat reminiscent of the position taken by those inflated masculine medieval theologians who granted a soul to men but denied it to women."

Even though in this paper I have mentioned Brunswik in a footnote and in a passing manner I hope the reader follows the maxims: “the devil is in the details” and “God is in the details.”

So what then is a cognitive system? To answer this I mashed a framework using my belief system along with my interpretations of the work presented by early cyberneticists, particularly that of Claud E. Shannon and Warren Weaver. Time and again, the words information and information processing occur in literature concerning cognition and I find it peculiar that researchers, especially speech community historians who call themselves linguists, don’t read up on information theory (Shanon, 1948) [6]. The framework concerning cognition and cognitive systems I’m using these days is explained as follows.

We start with an asseveration – information is ontological. Following this asseveration is the dogmatic statement that every physicality i.e., physical conformation, has the capacity to store information. What then could be the minimum amount of information stored in a physicality? The minimum information stored in a physicality is about itself. If the previous statement is not veridical then how would any observer ever manage to identify a physicality? Moving along this line of thinking, storage of information is termed as memory. Having memory is termed as knowing. Thereby every physicality knows, at least, about itself.

So, yes, the universe as a whole is thinking and it knows, at least, about itself i.e. every physicality it contains. And yes, it knows of all physicalities it contains while all those physicalities are also thinking… at least about themselves. If we are thinking about the universe while it is thinking about us, then what? The unending recursion in the answer to that question might not be evident to some readers so rather than attempting to articulate that answer let’s figure out an operational definition of a cognitive system.

A cognitive system is not one such minimalist thinker that has information strictly pertaining to itself. A cognitive system is that physicality which has the capacity to access, store and manipulate information concerning that which is other than self. It is a conglomeration of information that has the capacity to represent within its conformations other conglomerations of information. Also, it is able to do so without disintegrating and without turning into those other physicalities it acknowledges by representing them with full fidelity. A cognitive system is not a physical symbol manipulator it is an information manipulator and the rate of such manipulations is measured in bits per second. How come a cognitive system has such capacities? This question is the same as, “why does an observer have the capacity to observe?”

The current framework I have mashed together for myself doesn’t allow me to opine much on the subject of why an observer has the ability to observe anything in the first place. I see hope in being able to do so once I have a belief system that incorporates notions concerning “quantum decoherence” where I need to figure out if a waveform can observe itself and cause itself to undergo wavefunction collapse. I would also need to figure out how information “leaks” from one physicality into another. But until then I can continue to work at the engineering and managerial level where my chore is to simply rig physical conformations within tangible dimensions that channelize information between various cognitive systems. I can thereby help set up appropriate kan-ban within the system. This is obviously done by being a part of the systems for there is no way to exclude the observer from what is being observed.

I would say that the above notions concerning rationality and criteria for scientific exploration is what Karl Popper was also working on, in his own unyielding way. I think he wanted to find those intuited limits within which information about the universe can be rigorously corroborated by various observers while having awareness of the region beyond the intuited limits. That speculated region beyond the veil is where the corroborated hypotheses could become amenable to falsification even if they weren’t actually falsified until that date of socially accepted inquiries. The region beyond a framework’s limit and often times even within it can be beyond our perceptual bubble. But what is the point of having such a vast framework that clubs all cognitive systems under one umbrella? The point is to figure out how a particular class of cognitive systems exit, what their structural and functional components are and establish ways of rigging communication channels between different cognitive systems so that information can be shared profitably. By starting under the vast umbrella of cognitive systems one can later establish which of these is a table or chair and which is a human and figure out if there can be a way in which the two can communicate. A table does tell us, for instance, of the types of molecules present in the universe for at least the ones observed to exist within the table happen to exist in the universe. However, for us to be able to extract that information from the table we would have to set up a means of transducing that information through an electron microscope or certain such arrangement. The rigging would make use of the concordance between the molecules and the microscope and then the concordance between the microscope and our visual system.

This framework also makes it possible to figure out the levels of functionality a human can have based on our physicality and experiences. This framework elaborates a perspective in industrial engineering and human factors study which states that no human is disabled, one merely has different levels of functionality and engineering interventions can be put in place to accommodate a certain level of functionality. This is why an understanding of affordances allows us to cope with our “cognitive luck”.

But to cope with cognitive luck one doesn’t need to think outside the box, one merely needs to find or invent a box that is worthy to think within even if the walls of such a box are porous or dynamically shifting. The framework I’ve elaborated on is one such box.

It is very important to understand that those who have been an entrant into the universe through a certain set of embodied situations may in fact be gnawing at a science that is different than the one already digested, or rather, partially assimilated and partially discarded by those who have been entered into the universe through different means. This notion does not impact the right of a person to keep gnawing at various concepts nor does it impact their right to wright in accordance to various speculatives and operatives.

By understanding how every cognitive system is engaged in some form of art, for all that is perceivable or conceivably by any being is art, we can understand which ones are engaged in science. That art which is concerned with precise conceptions is science. This means that even people who are given to thought experiments are doing science because they are able to visit their conceptualizations with precision even if there might not be any measurable level of accuracy of their percepts with a reality that anyone else perceives.

Even if we chose a more constricted framework like artificial intelligence research paradigm, rationality might be a concept closer to the upper-ontology in a semantic space, sharing a position similar to the concepts of “concept”, “thing” and “happening” where finding the negations of concept, thing and happening is not analytically feasible. Thus, the word irrational might only be a lay term used as a tag for atypical human behavior much the same way words like silly or stupid are used in English speaking communities.

It is certainly evident that even after thousands of years of existence, human understanding of human understanding has remained rather incomplete and we often find ourselves caught up in things that are beyond our capabilities. So would such a behavior be termed irrational? Indeed, even Dadaism is not irrational, its rationale is to implement an artistic expression that is atypical within the market of art expressions such that the public might assign the colloquial sense of irrationalism to it. Persons who label themselves as irrationalists merely intend to adopt the principle of FISH – frivolously implanted stupendous hyperboles. So yes, pocket calculators and even soap bubbles are scientists for they manage to fulfill their functions within certain frameworks by precisely reacting to certain situations by way of their embodiment. If this sentiment is not amenable to a reader, the reasons would be the ones as explained up until now. Any sentiment is rational given some or the other context and it is possible that we won’t manage to find that which is irrational using analytical means for we might be trying to analyse what analysis is, without adequate contextual constraints. This is just an honest mistake that we all can make, it isn’t irrationality.

And actually, in my experience, irrationalists haven’t been the most peculiar lot amongst mankind. The most peculiar ones I encountered have been a group of people who happen to be “on spec”. These people appear to be searching for something without charter or warrant and when you ask them what they are looking for, they say, “I’ll know when I find it.” If you ask them whether they have any sort of criteria for identifying what they are looking for, they say, “The criteria will emerge spontaneously as that which is being sought gets found.”

Strangely, after encountering people who have been on spec I found myself to be on spec.

And all though I’d like to ramble on about rationality and science for even rambling or spinning yarns about anything while philosophizing can’t be irrational… I am tired.

References:

[1] Gigerenzer, G. (2006). Bounded and Rational, Contemporary Debates in Cognitive Science. Blackwell Publishing: MA, USA.

[2] Matheson, D. (2006). Bounded Rationality and the Enlightenment Picture of Cognitive Virtue, Contemporary Debates in Cognitive Science. Blackwell Publishing: MA, USA.

[3] Gigerenzer, Gerd; Selten, Reinhard (2002). Bounded Rationality: The Adaptive Toolbox, page 4. MIT Press.

[4] Kenneth R. Hammond’s essay Ecological Validity: Then and Now. See his essay for a review of Egon Brunswik’s concepts called representative design and ecological validity as well as the Brunswik Lens Model of cue utilization. Available online at: http://www.albany.edu/cpr/brunswik/notes/essay2.html

[5] Gigerenzer, G., Todd, P. M. & ABC Research Group, (1999). Simple Heuristics That Make Us Smart, Oxford University Press: NY, USA

[6] Shannon, C.E. (1948), “A Mathematical Theory of Communication”, Bell System Technical Journal, 27, pp. 379–423 & 623–656, July & October, 1948.

Advertisements

2 thoughts on “Rationalizing About Rationality: Ecological Rationality

  1. I just want to say I’m all new to blogs and seriously liked your page. Likely I’m going to bookmark your blog post . You definitely come with superb writings. Regards for sharing your webpage.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s