• Statistical Scientist Meets a Philosopher of Science: A Conversation
• Statistical Science and Philosophy of Science: Where Do/Should They Meet in 2011 (and Beyond)?
• You May Believe You Are a Bayesian But You Are Probably Wrong
• Induction and Deduction in Bayesian Data Analysis
• The Renegade Subjectivist: Jose Bernardo’s Objective Bayesianism
• Foundational Issues in Statistical Modeling: Statistical Model Specification and Validation
• Empirical Economic Model Discovery and Theory Evaluation
• Low Assumptions, High Dimensions

My (newest) position concerning causality is one that I’m fairly confident Pearl shares, which is good because, for one, it’s an unpopular position and, two, it’s always nice to have friends. Lately, I’ve been entertaining the notion of an “epistemic” conception of causality, which means that causation need not be thought of so much as a feature of the world but as a feature of the way we interpret the world. These are very different interpretations of causation since one implies that causal relations are things in the world (whether we observe them or not) and the other implies that causal relations are items of our understanding of the world (which may not be there if we’re not here to observe them). I think Pearl holds this position and I’ll give you some reasons to back up my claim.

First, when discussing the problem Bertrand Russell has regarding the problem of causation, Pearl is quick to point out that the asymmetry of causal hypotheses is not due to the structure of the universe as such, but is instead due to the ways in which researchers go about studying some small part of any physical system. Pearl says:

The lesson is that it is the way we carve up the universe that determines the directionality we associate with cause and effect. Such carving is tacitly assumed in every scientific investigation (p.420).

Now, while not a definitive acceptance of this epistemic conception of causality, it is still telling, I think. Pearl is responding to Russell’s criticism that in advanced physics causal language never comes up by arguing that the language of causality only comes into the picture when researchers focus on areas of the physical universe that are not maximally inclusive. In other words, it is only in isolated subsystems that causality appears: take a subset of the interrelations between physical objects and you will find causal relations; take the system as a whole and those causal relations disappear and the system seems symmetric.

There is something very interesting going on here with the related notions of “boundary conditions,” “input/output relations,” and “background conditions.” When we first isolate some subsystem and then characterize it in terms of inputs and outputs we draw a boundary around some phenomenon and, in so doing, create this asymmetry between what was designated as the input and the output. This idea is worth exploring further and, I’m sure, will become a much-debated item of discussion on this blog. It’s also in my dissertation, so there’s that.

Now, to be fair to the realist and a possible realist-reading of Pearl, declaring something to be an input does not mean that one must be an anti-realist regarding the nature of that element (inputs can be just as real as anything, after all). Now whether Pearl is or is not a realist about causation is going to be one topic of this multi-part discussion, but I am going to try to defend this rather extreme position I’ve outlined above and argue that, in fact, it is the labeling of certain phenomena as “input/output relations” that provides the directionality and the asymmetry of our causal hypotheses about those phenomena. So I agree with Pearl when he says that it is the act of investigation of an aspect of the universe that brings about causal language and I’m going to take him literally and defend the position that causation is just that; namely, the way in which we characterize systems in terms of asymmetric relations between elements of that system.

Like I said above, this is a highly unpopular and idiosyncratic position to hold, so I’m very interested in hearing your views.

Yes, that’s billiard balls. If you know your David Hume then you’re at least making some sort of face at my lame attempt to be clever…

Sadly, however, I have many points of dispute with the piece, including:

(1) The title is just silly, right? How is it a failure of “Science” if many (or even all) of our causal hypotheses turn out to be wrong? Is this a complaint that “science is hard”? Because that’s certainly true, but let’s not forget that it’s the same set of methodological practices that allow one to empirically reject bad causal hypotheses that allowed one to initially formulate those hypotheses in the first place. So, contra the author, it was the failure of those hypotheses (those reductionist hypotheses, note) and not a failure of science that is to be blamed.

(2) The author seems to burden “Science” with a particular interpretation of causation. But there’s two problems here; first, it’s a narrowly reductionistic and linear-style interpretation of causality he/she has assumed applies to the whole realm of scientific research; second, causal talk has not really been welcome in fundamental physics for around 100-150 years. Now, I disagree with the absence of causal talk in physics for many reasons, but it’s a gross error to generalize across disciplines and assert that this interpretation of causation holds across them all.

(3) This narrow interpretation of science, what the author rightly calls “reductionistic” is one and only one of many possible ways of conceiving of causality (and for discovering it via causal hypotheses testing). There are many proponents of causation in science who would flat-out deny that it works in the way suggested by reductionist approaches.

Anyway, feel free to read the article and argue with me.

Now the other bit of news. A few of us here at The Modern Dilettante have decided to initiate a new element to the blog; namely, a critical commentary on a series of texts, the first of which is called Causality: Models, Reasoning and Inference. We’re hoping to begin this commentary in the next week, so feel free to chime in or just follow along. We will try to keep the more technical and mathematical elements to the bare minimum and will focus instead on providing a clear and concise commentary (but it is a math book, after all).

But here’s a question. Since thinking critically about causation began with Hume…

…should we start there with an introduction to his thinking on the matter? It’s still highly relevant and every serious treatment of causality mentions his name in the first or second sentence of their work.

Reblogged from EvoAnth:

Ordinarily I give posts from the Institute of Creation Research (ICR) a wide birth because there is already a fine blog dedicated to commenting on what they have to say. Their latest post, however, requires comment. Mostly because they mentioned evolutionary anthropology by name (and as an egotistical male, I take any critique of my subject as a direct attack on me). But before I defend myself, some background. Göbekli Tepe is the oldest religious structure found so far, consisting of massive stone …

Interesting read

Here’s a picture of Darwin’s notebook, the page where he first conceived of his “Tree of Life,” taken along with him on his HSS Beagle journey, and which displays the dynamics of his thought in action. Note the “I think” in the corner. Amazing! I got to see this in person once and it was a nice and humbling experience.

But what is incontestably amazing is the simplicity of Darwin’s central argument. I’ll sketch it out and then provide a more rigorous treatment. Here’s the sketch:

As many more individuals of each species are born than can possibly survive; and as, consequently, there is a frequently recurring struggle for existence, it follows that any being, if it vary however slightly in any manner profitable to itself, under the complex and sometimes varying conditions of life, will have a better chance of surviving, and thus be naturally selected. From the strong principle of inheritance, any selected variety will tend to propagate its new and modified form (1859; p.5).

Nice, huh? Note how un-teleological (purpose-driven) the process is; many, many people, evolutionists included, fail to fully get that. Here’s the more rigorous argument (where P stands for “premise” and C for “conclusion”):

P1: That there exists an appreciable amount of variation of characteristics between organisms.
P2: That there is some process by which traits pass from generation to generation (inheritance).
P3: That individual organisms reproduce many more offspring than can be sustained within any given environment.
C1: Therefore, organisms will compete for a set of limited resources (from P3).
C2: Given C1 and P1, Some variations will be more beneficial to individual organisms who possess them.
C3: Given C2, individuals possessing “profitable variations” will have a greater capacity to survive that competition for limited resources.
C4: Given C2 and C3, P1 and P3, these individual organisms will also possess a greater likelihood for producing more offspring than competitors.
C5: Given the above, individual organisms with “profitable variations” will be more likely to produce offspring and, given that there is some mechanism of inheritance, these “profitable variations” stand a better chance of being passed in to future generations. In other words, that certain variations will be naturally selected and passed to future offspring.

Now, note that Darwin had no idea about Mendelian genetics. Mendel was doing his work a few decades after the publication of The Origin and was, alas, in German, so Darwin never read it. But that’s irrelevant since the argument only needs “some” mechanism of inheritance in order to be successful. The inheritance problem bothered Darwin and bothered him more when G. Mivart (I think) leveled the “blending inheritance” problem at Darwin and his followers, but that problem was solved by Mendelian genetics (but not by Mendel himself) and the Modern Synthesis of Fisher, Haldane, and Wright (I’ll post on that later…).

But just appreciate the utter simplicity of Darwin’s argument and note that this structure works for any type of selection..for any set of organisms anywhere and anywhen, so long as those meager premises are fulfilled! And people are impressed by Einstein’s equation for its generality…

I wanted to point out two things in this post. First, the major climate models in use today are open source; i.e., the code for the mathematical models are freely available to the public. Check out, for example, the mathematical model developed for NCAR called the CESM (community earth system model), which is freely available along with the relevant dataset. Interestingly, that seems to put climate change antagonists in an “uncomfortable” position since, for one, all of the relevant material for making an informed assessment of the science is right there for the perusing.

The second thing I wanted to point out is that most of the climate models that are in use today (which are ridiculously mathematically complex) generate robust results. What this means is that the models are producing exact or very similar predictions for numerous iterations of the model. Now what makes that “reallllly” interesting is that the parameters and values for the multitude of variables that appear in the models are different!

Let that sink in for a second….

Why would different and mathematically distinct models of some phenomenon ever be expected to generate the same predictions when they have different parameter values etc? These models share “a common causal core” which is, I’m sure you’ve guessed it, the Greenhouse Gas hypothesis. So you dump this hypothesis (that greenhouse gases possess such-and-such chemical and physical properties) into a ton of very different models and you get the same result; e.g., that global warming is a reality and that we are most likely the cause of it.

Coincidence? Hardly. Discuss.

Few continental philosophers ever took up Husserl’s notion of philosophy as a rigorous science.  And even fewer analytic philosophers, since it was William James that was responsible for Husserl’s work not being published until 60 years later into English.  While it remains quite under-studied and certainly controversial, whether one adopts the phenomenological nature of philosophy as a universal science, it opens up a question popular to all three philosophical movements at the turn of the 20^th century.  Peirce as the founder of American pragmatism, the analytics of the Vienna Circle, and Edmund Husserl founder of continental phenomenology, were all committed to the relationship of philosophy and science, and to saving philosophy from Hegelian metaphysics.  What Husserl invites us to look at anew is the very real possibility that philosophy could be a universal science in a way that the positive sciences cannot because of the difference in their subject matter, philosophy being an anti-naturalist science of logical inquiry, of analysis, of pure description, of sign relations, of meaning, of essences.

Its been 101 years since Husserl wrote the article. Now going into the twenty first century, I wonder whether this project is the right course?  How can we look at Peirce, or Husserl, or the logical positivists and demarcate an objective methodology that advances philosophy beyond the antiquated metaphysical projects of systematic philosophy?

Now the idea Husserl had in mind was that we could have a systematic philosophy that was not scholastic metaphysics or Hegelian dialectical metaphysics or anything of the kind.  Instead, Husserl believed philosophy must be radicalized into a science of reason, a science of consciousness, a universal science of the relationship of logic and the pure idea.  The word idea, of course, is a loaded term that makes philosophers almost as nervous as the word intuition (or maybe that’s just me?), so I will take this time to elaborate before a misunderstanding of connotation arises:

“If all given theoretic unity is in essence a unity of meaning, and if logic is the science of theoretic unity in general, then logic evidently is the science of meaning as such.” (Logical Investigations, 225)

“This notion of logic as a science of meanings is of course at odds with the mode of speech and treatment of the traditional logic, which operates with psychological or psychological slanted terms such as ‘idea’, ‘judgment’, ‘affirmation’, ‘denial’, etc., which thinks it is really only establishing differences of psychology and tracking down psychological laws.” (Logical Investigations, 225)

Idea is not a feeling or a mental picture in a person’s mind nor is it a behavior-stimuli from the ontogenesis of learning a word or an object.   Contrarily, Husserl conceives of a logic as a science of pure description.  However, in order to accomplish this epistemologically, the primary criteria of philosophy must be based on evidence (the noema), whereby the conditions for fulfilling a state of affairs must be examined and demonstrated to explicate the truth of something.  For this to be achieved, the separation between being and consciousness (subject and object) must be “subjectivized” by using a method of logical analysis to take all the subjective content of experience, the things-that-appear (vs. the thing-in-itself) and make the content of experience the object of scientific, rational investigation.  As Quentin Lauer puts it, the logical nature of experience would be treated as the investigation based on the nature of the content, “whether that be an object, a state of affairs, a process, an event, a social reality, a culture, etc.” (Phenomenology and the Crisis of Philosophy, 45).

From Descartes to Hegel, systematic philosophy was primarily an epistemological project determined to unify knowledge upon a theoretical foundation of first principles (empirical, rational, or critical) with the end purpose being the ultimate metaphysical unification of the Cartesian problem (or not, if skeptical) of subjective knowledge with its dualistically corresponding objective mind-independent essence.  Twentieth century philosophy, however, can be interpreted largely as a reaction against this more or less uninterrupted four-century-old project on several levels (not going to get into the linguistic turn here).  Most importantly, philosophers began to question the methodology underlying systematic philosophy as too scholastic in principle, establishing a finite set of second order concepts or principles as constructing a purportedly comprehensive determination of the nature and structure of knowledge and reality around its foundational presuppositions. As the neo-Kantians discovered with the architectonic categories, there is often a sense of arbitrariness or at the very least a lack of objective methodological criteria one can demonstratively give through reason alone for justifying their philosophical system for conceptually interpreting scientific reality.

Edmund Husserl believed philosophy could not progress until it became science

Such seemingly endless nineteenth century neo-Kantian debates throughout Europe defined the intellectual world Edmund Husserl found himself as a mathematician turned philosopher.  It may surprise people, however, that Husserl was philosophically most influenced by the epistemological challenge of Descartes and the rigorous skeptical empiricism of Hume.  It was in the spirit of the theory of knowledge that Husserl came to believe that Western philosophy was stuck on a treadmill, unable to progress forward.  Husserl came to see modern post-Cartesian philosophy as a series of individualistic philosophies without a clear, objective procedure for analyzing reason.  If reason was the ultimate domain of the philosopher, why was there not an objective scientific philosophy that demarcated the applications of reason as a tool and the limits of it as a method of investigation?  How was it that philosophers could simply take a leap of faith that reason was the philosopher’s instrument without agreeing to how that instrument instructed our discipline between those that rely upon it?  Indeed, Husserl was concerned that reason itself had never been properly investigated in a rigorous way.

This led Husserl to his Logical Investigations, which was in part inspired by Frege’s critique of Husserl’s Psychology of Arithmetic.  Husserl reconsiders the nature of reason as an objective logic whose truth and procedure is ultimately mind-independent of naturalist and historical facticity.  Contrarily, Husserl investigates the fields of mathematics and logic as transcendental sciences whose truth is justified by evidence other than factors reducible to empirical considerations.  What about other fields of knowledge?  Here we see the central focus of Husserl’s life is the question of epistemological evidence, of meaning-intention and meaning-fulfillment of appearance (see noema vs. noesis).

Hume provided the greatest modern challenge to the notion of universal science (or science in general).  The idea was simple.  Taking the dichotomy of matters of facts and relations of ideas (Leibniz’s truths of fact vs. truths of reason) Hume stipulates that the only knowledge one can have is through cognitive analysis.  However, as Hume and Kant understood, this adds no new knowledge.  This is the old paradox of conceptual analysis, and why Kant was obsessed with the synthetic a priori as the only means by which to add new knowledge without falling victim to Hume’s devastating skeptical critique of factual, contingent knowledge.

Like every serious epistemologist, Husserl took Hume as the central challenge of scientific possibility.  However, Husserl had a dilemma.  At the time, the neo-Kantians were at the height of their popularity and Husserl, to put it nicely, thought Hegel was a dead end with no program to advance.  However, the problem with Kant was the ding an sich, the distinction of phenomena and noumena did not allow an easy solution to bridging them; and Husserl did not believe the answer was waiting in the synthetic a priori.  The problem for Husserl was that Kantian philosophy transcendentaized the form of experience but left the content of experience forever suspect.  For Husserl, the Copernican Turn never finished advancing, Kant was never radical enough, as Kant failed to obtain the transcendental science he spoke of in the Critique of Pure Reason but never developed in his own work; the problem was content of experience needed to be radicalized in the Copernican Turn too.