Before I left I was in the process of writing a series of essays on “Marxism as a Science.” The early essays are “infra-philosophical” in the sense that they deal with mixed philosophical material: the material exigencies of sustaining a school of thought combined with the conceptual “fitness” of Marxist ideas. Initially, the idea of “window dressing” and retconning Marxist theory to look scientific, which I took to be represented by Michael Burawoy’s work, I found repulsive because of an ingrained philosophical tendency to reject “respectability politics.” However, the severe problems of the social sciences and Marxism were very quickly revealed to me. In short, 3 essays later (and I was in the process of writing a 5th) I now see Analytic Marxism (AM) and Rational Choice Theory Marxism to, in many ways, be the best thing that ever happened to Marxist theory.
Its well documented failure is often attributed to its hermeticism, theoreticism, and separation from actual politics. While these factors certainly played a role, I see AM’s failure as being more to do with its failure to build an institution like the ONPHI. AM was a pre-internet phenomenon that was thoroughly caught up in the university and made no pretense to build an institution outside of the usual channels of knowledge dissemination (university publications, academic journals, lectures, etc.). I won’t speculate further on the cause for AM’s failure except that I find Andrew Levine’s argument in A Future for Marxism? Althusser, the Analytical Turn and the Revival of Socialist Theory that “it was philosophy, more than anything else, that did Marxism in” very compelling.
This is what brings me to the notion that non-Marxism is a potent “assistant” to Marxism insofar as it strives to be a science. I wrote a while back:
At least a few courses of action present themselves, some of which we have already discussed: 1) academic Marxists henceforth become “data-driven” economists practiced in the intricacies of policy writing, 2) Marxists rebrand by either totally disappearing and shedding the proper name “Marxist” or “Marxian,” or 3) Marxists reimagine their discipline as a unified field and begin to aggregate and assimilate non-Marxist knowledge (the specter of dialectics) for the sake of making Marxist arguments. However, while we have already been acquainted with the non-philosophical field, Laruelle’s proposal for a “non-Marxism” or “non-standard” Marxism has not been addressed. The wager here, which is an occasion for a “report” on non-Marxism, is that these courses of action are best “in-formed” by the study of non-Marxism, that the pursuit of science by the Marxist is inevitably still too enthusiastic, still too concerned with theoretical “correctness,” and undertaken as a matter of duty or obligation which would imply a symptomatic and self-aggrandizing “suicidal narcissism” which, simply put, is unsustainable. This is what Laruelle derogatorily calls the Maoist “long march.”
This “work” must not be undertaken to effectuate the equation Marxism=Science with the knowledge that Marxism has scientific pretenses and aspirations which it deserves to achieve for its own sake so that it may reestablish its authority and sufficiency: this would be the Marxist “conquest” of science which would just be a dangerous and reified variation on philosophy’s confusion of materialism with science, i.e., materialism=(the) (essence) (of) science so that sciencem > sciencei (where “m” and “i” represent “materialist” and “idealist”). There can be no “subordination” of science to any particular conceptualizing philosophical or political stricture. What is meant by this is not the idealism that it “knows best” or will “find a way” (à la Jeff Goldblum in Jurassic Park’s statement “nature always finds a way” presumably out of this or that limitation by its adaptability and innate drive to “be free”), but rather that science will always continue since actual scientific research is a differential procedure and always conducted (and succeeds or fails, advances or stagnates) according to its immanent exigencies. Which is to say, according-to-the-Real of whatever it is trying to model and according-to-the-One of its political, technological, political exigencies (—if this is odd, consider how “science” was undertaken in the 1600’s and how it was considered science in accordance to its own principles). What this means is that the enthusiasm and enlightenment of the dialectically thinking Marxist scientist really amounts to a set of subjective priors used for inferential reasoning that provide no significant cognitive advantage in scientific research and may even be disadvantageous. Meanwhile, when it comes to the conceptualization of scientific research, Marxist or dialectical thinking usually amounts to the felicitous narrativization of scientific discovery as an advertisement for Marxist theoreticism or an occasion for tired polemics.
This is to say, it must not evaluate non-Marxism (or any field for that matter) and then take what is assimilable or amenable to it—this would be the erection of a specular ideal-ego mirror which would simulate science while still reducing theory to matters of orthodoxy, taste, and political jouissance. Rather, Marxism must strive to become non-Marxism or realize itself as having always already been non-Marxism, which is to say, it must abandon its integral/integrating tendencies and end its dialectical conjugations and translations of non-Marxist or “not-Marxist” (non-)knowledge and adopt a quantum or algebraic approach to knowledge(s). It must think in terms of superposition and complementarity rather than the (previously established) philosophical “arrangement” and “total” presentation of knowledge according to quasi-hermeneutically defined “limits,” “gaps,” and “contradictions” which are historically-philosophically narrativized. It is true, all knowledge “belongs” to Marxism insofar as it is the “science of the proletariat,” or a form of non-philosophical “gnosis,” but knowledge cannot be subordinated to its authority or the authority of its exponent theorists and their univocal interpretations (knowledge policing): its authority must be effectuated or effectuational and established legislatively and consensually as democratic and pluralist.
It was already clear to the analytic Marxists that updating their field would require abandoning much of its identifying traits; however, to undertake Marxism becoming non-Marxist is to yet further purify Marxism of itself and to render it almost generic, algebraic and variable, such that (non-)Marxism=X, while at the same time “killing ones darlings” in an unprecedented manner (e.g. the painful and so far incomplete abandonment of the LTV or the disproving and final relinquishment of Marx’s transformation problem). The shift to quantitative knowledge production has already been made by [some elements of] Marxism and is being made by the broad Left in general—this is inevitable. However, if the theoretical and political elements of Marxism are to become scientific according to an un- or non-reified notion of science, that is to say, non-philosophical (rather than just anti-philosophical, which is always still philosophical), they must engage science, non-philosophy, and non-Marxism in accordance with their own standards of rigor.
If I remember where I was going with this, I was addressing the recent interest in data-journalism by publications like The Jacobin, Salon, and Mother Jones’s new “wonkish” statistical approach to some of their reporting following the trend set by Nate Silver’s (now somewhat crappy and frivolous) 538 blog. I now regret spending time on Silver’s book in my large neurotic 4th essay on Marxism and Science. I found a pool of very interesting french websites on data-journalism where the discussion is more lucid. When I have the heart I might revamp that section of my 4th essay with those sources.
Anyways, I don’t know if Mother Jones counts as a “the broad Left in general” but I’m sure I was getting at Picketty’s data-driven “intervention” into the contemporary political discourse on equality (taken up by my state senator, Liz Warren). Progressive does not equal “broad Left,” from my scan of a massive amount of Marxist blogs last year in my bizarre essay on Žižek and SYRIZA the arguments remain embarrassingly “dialectical.” Maybe this is wishful thinking. However, an example of “not-Marxism” being used in the service of Marxism is the AM group’s use of methodological individualism, rational choice theory, and neo-classical economics. However, these “cutting edge” practices were considered the standard for what passed as contemporary conventional science, which is a bizarre evaluation. The “enthusiasm” of the shift cannot be discounted. We can see this in the two definitions but forth by the AM and RCT Marxists (citations below are from “Review Essay: A Future for (Analytical) Marxism?” by Roberto Veneziani):
Definition 1. Analytical Marxism (AM) is defined by an analysis of Marxist concerns that is focused through
C1. ‘A commitment to conventional scientific norms in the elaboration of theory and the conduct of research’.
C2. ‘An emphasis on the importance of systematic conceptualisation [ . . . ]. This involves careful attention to both definitions of concepts and the logical coherence of interconnected concepts’.
C3. ‘A concern with a relatively fine-grained specification of the steps in the theoretical arguments linking concepts’.
C4. ‘The importance accorded to the intentional action of individuals’.
Definition 2. Rational Choice Marxism (RCM) is defined by an analysis of Marxist concerns that is focused through C2, C3 and
C1’: The use of ‘state of the arts methods of analytical philosophy and “positivist” social science’
(Roemer, 1986c, pp. 3–4);
C4’(i): MI, ‘the doctrine that all social phenomena – their structure and their change – are in
principle explicable in ways that only involve individuals – their properties, their goals, their beliefs and their actions’ (Elster, 1985, p. 5);
C4’(ii): Rational choice explanations. This ‘involves showing that the action was rational and that it was performed because it was rational. That the action was rational means that given the beliefs of the agent, the action was the best way for him to realize his plans or desires. Hence, rationality goes along with some form of maximizing behaviour’ (Elster, 1985, p. 9).
The drive towards “coherence” ought to be celebrated. I feel that non-philosophy is somewhat weak here in its propensity to engage in typically French rambling that can bury the thread of the argument for anglophones (its evident that French loses a lot of its specificity when it is translated into English since it is deprived of its celebrated “accord” between gender and number which adds a lot of clarity). However, this seems way too keen to drink the rational choice theory kool-aid for the sake of hysterical conservatism rather than practicality and when the critiques of RCT and MI were even in 1989 well known, especially in left-wing circles. Inventing new methods or harnessing RCT and MI without adhering to any of their assumptions would have probably been preferable. Imagining a RCT without rationality is something I have considered elsewhere. The choice between hysterical conservatism or “indulgence” and domestication of opponent theories seems too limited, non-philosophy and non-Marxism can easily bypass this conceptual deadlock through their “non-standard” or even “queer” conceptualizations of apparently conceptually homogeneous fields of study and their attendant practices (e.g. psychoanalysis and non-psychoanalysis).
Thus we must reverse a major thesis of the previous essay which was still too Badiouan, still too philosophically committed and encumbered, and still too infatuated with the brand “Marxism” and the proper name “Marx” to effectuate a “Gestalt shift” away from dialectical materialism: Marxist thought should not strive to “compossibilize” knowledge so that it presents potentially contradictory and controversial knowledge as a way of outmaneuvering it and domesticating it through the policing of its dissemination via its usual pedagogical procedure (“Marxifying” this or that knowledge for the sake of the always infantilized proletarians). Further, the bioethical and ecological prerogatives of Marxism (previously “conjugated” or “transposed” with this bioethical emphasis), while they are useful as a way of producing a provisional “One” (Vision-in-One) with which Marxism can “scientifically” work “in accordance to” or be “determined” by as a regulative ideal, can only provide a limited conceptual framework or metaphorical/descriptive infrastructure. This rather than a fully blown epistemology with all their usual pretenses for a foundation of knowledge and truth let alone a substantive scientific practice.
In quasi-Deleuzian or Baudrillardian terms, insofar as capitalism remains thought within the confines of philosophy and capitalism and political economy are fraught, confused with, integrated with, etc. their own (mis)representations and conceptualizations, the reality of capitalism will always be a foil to its Marxist conceptualization via its simulacral and drive-like inertial desire to model itself and anticipate its own movements; this is a simple impossibility since it’s complete integral modeling would just reproduce the complexity and mystery of capitalism’s aggregated causes and effects (i.e., its uncertainty principle). Capitalism is always-already compromised yet reproduced by its predictive self-modeling and self-simulating activity and (i.e., its complementarity or quantum indeterminacy). In more simple non-philosophical terms: capitalism, the economy, the market, and so forth are all foreclosed to thought while remaining eminently conceptualizable and philosophizable: this non-relationship makes possible the very theoretical impetus that enacts and discovers this impossibility in the same way that for Lacan the non-relationship between man and woman produces the (im)possibilizing fantasy of the sexual relationship. Insofar as Marxism is inevitably drawn into and apart of this quagmire, it would seem that the proper metaphor and conceptual neutralization/expansion of Marx’s original scientific naturalism and materialist dialectics is that of a network comprised of reciprocally determined actors (Latour): hence our major thesis is that actor-network theory must replace dialectics.
I prepared notes on a presentation by Latour on actor-network theory but they are too unorganized to present here and never made it into my essay. Apparently this idea has already been proposed by Speculative Heresy and mulled over by this fellow shiviro. I’ll follow this path and pick up Prince of Networks by Graham Harman at some point.
What later follows is my definition of “Marxism-as-a-Science” which deflates (or inoperatizes) truth claims by reconjugating them as propositions that must be statistically testable not just analytically valid. This follows from the discourse I was having on econometrics with Graham Joncas. I want to stick with econometrics and statistics a lot more before venturing down the network actor theory path for “strategic” reasons. Econometrics remains a much more potent (read: respectable) and delineated practice with a field of well understood problems. Actor-network theory (ANT) and scientometrics remain somewhat caught up in epistemological and typically post-structuralist theoretical concerns. There is a certain rigor and practicality to econometrics and statistics that I appreciate. I highly recommend Writing Capital’s essay “there is no economic world.” and his piece on non-knowledge on this subject.
Marxism as a science is a decisive formulation since it demands that Marxism be thought according-to-science or determined in-the-last-instance by science, hereby Marxism-as-a-science gains its positive characteristics separately from Marxism and only insofar-as-it-is-a-science. Marxism-as-a-science must be rigorously differentiated from Marxism and, at least in this instance, from non-Marxism since the “non-” of non-Marxism is indicative of its heretical universalization/generalization according to its non-philosophical DLI. Marxism-as-a-science cannot be generalized and is non-generalizable although it produces material that is generic and which can be treated by non-philosophical procedures (e.g. forces of production, labor power). Marxism-as-a-science, since it is already purged of its pretention to be a semblance or mirror of science insofar as it is conceptualized, already establishes itself as a non-standard and non-philosophical Marxism. It uses the same material as non-Marxism and undertakes some of its same procedures, however, non-Marxism is algebraic and generic, it uses dualysis and cloning as its procedures (such that Marxism=X) so that it may analytically parse Marxism’s philosophical and transcendental material according-to-the-One or its DLI.
Marxism-as-a-science, however, has a more obvious affinity with the social sciences and is characteristically statistical rather than algebraic: it attempts to think quantitatively and according to the immanent exigencies of statistical science, i.e., according to statistical models which it takes provisionally as instances of structured data which have the totally variable status of “the Real-according-to-data” and the dilemmas produced by statistical inference. It aspires to adequation-without-correspondence and attempts to non-decisionally gauge “effect size” and “impact” or (statistical) “significance” in accordance to given instances of structured data. Its work of interpretation is inferential and probabilistic rather than non-philosophical, i.e., quasi-hermeneutic or anti-hermeneutic (cloning, dualysis). It does not take a specific science as its “inspiration,” supplement, or conceptual infrastructure, which is to say, a readymade field of terms that can be “deployed” for non-philosophical usage (e.g. Laruelle’s use of quantum physics); rather it takes material from the fields that most immanently confront it (sociology, economics, political science) with no pretense to fictionalize them in a philo-fiction, operationalize them, or turn them into material for non-philosophy though it readily makes use of non-philosophy and non-philosophical material (its methods, language, and “force-(of)-thought”) itself: the “writing” of Marxism-as-a-science is never Marxism-as-a-science at work, it is only the non-philosophical (in the sense of not-philosophy) and inevitably insufficient conceptualization of Marxism-as-a-science. Marxism-as-a-science has little anxiety about philosophizing, its philosophizability, or its being philosophized about insofar as it is “just,” which is to say, non-decisional and done according to or insofar as Marxism-(i/as)-(a)-science, which is to say, that it relies on philosophy in the same way that science does, i.e., as a translation, preservation, and historicization of its work in the form of conceptual knowledge so that it can (re)discover its gaps, its priorities, direction, and long term trends as an evening thought that might in-form its practice.
There is no outmaneuvering Laruelle in the following regard, we have already comprehended that Marxism can no longer be “intensified” or “supplemented” by anything–that to hold on to the Marxist corpus is to board a sinking ship, that any future Marxism must be radicalizing and irreverent, which is to say, heretical. The Marxist corpus must become a “corpora” for a kind of non-Marxist/Bayesian “machine learning” procedure that reduces philosophical Marxist phrases into material to be dualyized and philosophically deactivated, i.e., generalized, and underdetermined (much in the manner of the semantic procedures of the analytic Marxists: transforming Marx’s work into valid propositional statements). Meanwhile, Marxist economic statements must be turned into “incidences” that can assist Marxist philosophical claims. In other words, Marxism-as-a-science must produce the meta-analyses and interdisciplinary non-experimental longitudinal (historical) studies of capitalism that it already does but as an “econo-fiction.” The division of labor is thus: the former philosophical material (its philosophical uttrances) functions as a scientific provocation that must guide a kind of translation into a parametric statistical sociological or econometric model, the latter economic statements function as “disinterpretations” and “de-conceptualizations” of all philosophical statements about the economy or society: they are political neutral and regionally factual–in aggregate they produce the policy writing of the Marxist analysis and political perspective.
The scientific “force-(of)-thought” here would be to reduce the number of assumptions in the statistical models by establishing generalizable inferences as provisional axioms (e.g. TFRP); the analysis would aim to be non-parametric for the sake of lowering the “risk” of making Marxist claims in simple models while increasing the statistical power of more “risky” complex models (this is assuming the axioms are, in a mathematical sense, true). By this procedure the DLI becomes both more generalized and more robust in its capacity to “differentially” or “syzygistically” account for exogenous and endogenous variables (Joncas), however, in revenge the DLI loses its conceptual descriptive capacity to (philosophically) determine an efficient or final cause since cause and effect since, from the standpoint of statistics, all this disappears into noise, i.e., “effect sizes” (ES) and “statistical significance” and the ambiguity of p values. Insofar as the causality of traditional philosophical statements (which invariably take an if/then form) can be quantitatively modeled, they can only be accounted for according to conditional scenarioizations that are accounted for if certain “experimentally” defined (that is to say “knowledge based”) threshold conditions are satisfied (the “if”). The advantage is that this allows the model to account for “dynamism” or “complexity” via transitions to other “state spaces” or different “discrete-states” (the “then” might be accounted for by a DTMC, although there are other ways of modeling dynamic or complex systems) but at the expense of divining a specific cause since it disappears into a variable which represents a range of causes and conditions. In other words, the DLI loses the ability originally granted to it in historical materialism to transpose a conceptualization (explanation, description, and identification) of a cause, usually economic, into thought so that it functions as a rule or an axiom which can be used in a kind of mixed procedure of analytic deduction and forensics. This is to say, Marxism-as-a-science typically undoes the PST which supposes that thought can divine cause since it models reality statistically axiomatically treats it as quantitative in nature; it admits to the impossibility of making the simplifying assumption of the historical materialist DLI; it refuses to leave excess and captures unknown causes as “factors” accounted for in an exogenous variable; it refuses to conceptualize the totality or even n-number of specific efficient causes so that any “causal power” ocan be attached to a “term” as a property; it refuses any causality that is not relative to the One-Real. Rather, Marxism-as-a-Science quantizes the DLI so that it accords or abides to the Real quantitatively. The DLI is thereby sutured to a statistical model and arises “according-to-the-Real qua data,” i.e., a provisional and non-decided Real. To “determine,” hence forth, means to non-decisionally and non-philosophically determine according to the exigencies of the experimental data; hereby the philosophical “if/then” becomes a matter of establishing the correct procedure of statistical inference. The arbitrariness of the DLI is thereby expunged since any philosophical causality immanent to “terms” is removed and becomes a procedure of statistical inference and modeling. Though such terms are not necessarily illegal, they are given stringent criteria for adequately representing statistical models, are turned into probabilistic inferences, and are translated into “productions” in a dynamic/stochastic/knowledge-based logic model.
The immediate result of this is the disruption of hasty Marxist truth claims and the historical materialist DLI; however, surpassing the initial frustration of this disruption optimally leads one to the task of scientific work which, for a (non-)Marxist, aspires to a comprehensive meta-analysis. The necessarily mathematical and computational aspect of this work allows the Marxist researcher to finally benefit from the cooperative or collaborative, non-discursive, and democratic aspects of true scientific research: he or she is solving a problem, building a model, and proving a hypothesis–not trying to “phrase” evidence in such a way to support a claim. They have transformed from a scavenger of “stylized facts” which are then aggregated into a persuasive argument into a producer, which is to say, a worker. Finally, they are given immanent requirements for scientific rigor in the capacity of the proof of their hypothesis and maximization of the predictive capacities and “correctness” of their model and data structure.
Here the predictive power of the model is commensurate with its “truthiness.” Call this capitalist “truth” or a bankrupt form of bourgeoisie or utilitarian notion of “truth as what is useful” but should the model do what it claims it can do than it stands (the reverse of Marxist “theoretical” models “which always say what they do” and “never do what they say”). Meanwhile, any model can be expanded and complexified (albeit not without notable risks). The majority of the theoretical “hiccups” (that is to say, the kind that might bother Feyerabend and that obsessively want to be resolved by theorists like Campbell) that occur around model making are, as we identified in the last essay, around their description, their explanatory capacity, their ability to be translated into economic or social policy, and their instrumental or proprietary (non-scientific/biopolitical) use—they are otherwise fairly adequate in themselves as long as they have no pretense to model the “world” or become “worlds” in themselves. The difficulties provided by the problem of “incommensurability” at the level of the model are largely philosophical/ontological, i.e., a problem that is “solved” by the notion of “adequation-without-correspondence” while the most severe difficulties occur when natural language is introduced, in which case the most red-blooded scientist can become a sincere correlationist, an irresponsible post-modernist, or a confused and potentially non-committal realist or pragmatist—with little in-between but paradoxically a lot of entry room for spontaneous “transcendent” or “philo-scientific” ecstatic ideologies by virtue of science’s dual foreclosure to philosophy and philosophizability (the same goes for the economy). The solution to this seems stupidly disciplinarian: the fostering of a “relative competency” for the interpretation of statistical data (statistical cognition) and attitude of disruptive pan-critical rationalism about any and all interpretations: the inoperatization and censorship of non-rigorous interpretations, conceptualizations, generalizations, politicizations, or overall motivated and untactful “skewing” of statistical data.
Its going to take a bit to reboot this project. The major gaps are obvious: an analysis of Latour’s actor network theory, an analysis of the “progressive tendencies” of AM and the awaited publication of the translation of Laruelle’s book on Non-Marxism (although my French should be up to snuff to read it in the original language). A decent practical grasp of statistics and matrix algebra would be necessary for me to authoritatively continue down this line of reasoning (although I’ve been in dialogue with some people who do know and they say I’m not totally insane). This project might have to be abandoned or split into shorter summaries. My original plan was to include an section on Lyotard’s Libidinal Economy. I’ll have to see if this project sparks any interest.