Second Order Observation Historicized—An Epistemological Frame Narrative

Niklas Luhmann who, if he did not invent the concept of »second order observation,« gave it enough productive use to launch these three words into the international and interdisciplinary circuit of philosophical notions, never applied them in any historical description or in any historical analysis of real world phenomena. What would then be the point, what would be the intellectual bet implied in our complex experiment of historicizing the concept of »second order observation« and the act of observing oneself in the act of observation? Or, to ask the same question in an even more meticulous way: Why do we replace the concept of »self-reflection« with that of »second order observation,« and what do we expect from the historicization of the concept and of the phenomenon to which it is supposed to refer?

The claim is that such an operation may help us to understand, to explain, and to compare, in a more comprehensive fashion than before, certain stages of what we have come to call »the process of Modernity,« in particular the decades between 1780 and 1830 which we have long known as the age of »secularization,« of both »Classicism« and »Romanticism,« of the »crisis of representation,« according to Michel Foucault, and as Sattelzeit (»saddle time«), according to Reinhart Koselleck. Now, while it is plausible to say, in the first place, that the meaning of »secularization« does not cover many of the historical structures and events that we associate with the period in question and while the simultaneity and tension between Classicism and Romanticism have always puzzled cultural historians, it may well produce an impression of arrogance, especially in the contemporary academic-intellectual climate, not to follow such eminent and deservedly canonized thinkers of the past as Koselleck and Foucault.

My justification for this double omission, I hope and believe, is easy to understand. If we start with Koselleck’s notion of Sattelzeit, we will discover that its scope is exclusively hermeneutic. It states that texts from before 1780 will always require historical commentary because of their high degree of semantic otherness in relation to our own social knowledge, whereas texts from after 1830 regularly expose us to the risk of not taking their otherness as seriously as we should. But with the concept of Sattelzeit Koselleck does not try to point to any of those historical transformations that might explain the hermeneutic difference in which he is interested. By contrast, Foucault’s historical narrative about the »crisis of representation,« first developed in Les mots et les choses, his most important book, published in 1966, does describe a dense mood of skepticism and frustration in the world around 1800, a skepticism about the possibility of pinpointing, cognitively and conceptually, the phenomena of the world. Foucault believed that a new epistemological configuration in which »man« was simultaneously subject and object of understanding, together with a radical historicization of all phenomena, had emerged as a complex solution to this problem in the early nineteenth century. But while his analysis was truly innovative and still appears largely convincing today, he was not able to show why the then-new epistemological position of »man« and the surge of historicization occurred at the same time and how both movements may have complemented each other.

So much for our reasons for historicizing second order observation. Another question about the legitimacy of our project concerns the sheer possibility, rather than the pertinence, of historicizing second order observation. For it must be said that a certain structure of self-observation belongs to the core elements of human consciousness, is thereby temporally coextensive with Homo sapiens—and hence certainly not capable of being historicized. I am referring to the double structure in human consciousness that Edmund Husserl described with the concepts of »protention« and »retention,« that is to the capacity of each present moment in our mind to “resonate” with its previous moment and to »anticipate« its following moment and to thus always be, at the same time, inside (in its own present) and outside of itself (in the present’s immediate past or immediate future). While this structure may be specific to the human mind in comparison to animals (we will most likely never know), there indeed exists a different type of self-observation in the strict sense of the word that is even more likely to be exclusively human—and this is the ability to see our own minds’ function, an ability which Husserl’s phenomenology was relying upon as its elementary operation and as its basis of experience.

I do not believe, however, that Luhmann had one of these two modes of self-observation in mind when he spoke of »second order observation.« His notion was more hermeneutic, in the sense of seeing oneself »as if from outside,« as against the backdrop of a potential multiplicity or even a potential infinity of alternative forms of behavior at each given moment. Seeing oneself in the midst of possible alternatives is a way of constituting one’s self-image as a structure of meaning, and for that reason I call this type of self-observation »hermeneutic.« But how can this mode of self-observation be historicized? What would it have to do specifically with Modernity as a historical dimension? Weren’t the pre-Socratic philosophers, Socrates himself, or Michel de Montaigne, for example, outstanding second order observers in that very sense, outstanding observers of what was specific about themselves? The answer is that what can be and was indeed »historical« about this mode of second order observation during the historical period in question, and what can give a historical dimension to any type of human behavior, is simply its capacity of becoming habitual. In this sense, at least for a certain group of humans, second order observation stopped being a specific operation of the mind around 1800 and became habitual, that is it became part of a sociologically specific life form. In this very sense, too, it appears plausible to assume that during the Sattelzeit, self-observation in the now-specified meaning of the word became habitual for those individuals who liked to think of themselves as »philosophers« and whom, from around 1900 on and for more than a century now, we have been calling »intellectuals.« From a sociological point of view, intellectuals cannot help observing themselves in the act of observation; they are habitually self-reflexive, and this historical innovation in the shape of a certain social role, which occurred around 1800, has had considerable epistemological consequences. It is the intention of my essay to describe these consequences—and the historical consequences of these consequences—in the form of a frame narrative that may give orientation and shape to more detailed historical research in the future.


I will not try to come up with hypotheses about the »historical reasons« that led to the emergence of second order observation (as a habitual feature of intellectual life around 1800), for »historical reasons« on a similar level of complexity and abstraction tend to be tautological re-descriptions of the phenomena that they mean to explain. Would it really add anything, for example, to say that the emergence of second order observation was a reaction to and a means of coping with an ever-growing world complexity—except for stating that the world of the eighteenth century was likely more complex than that of the seventeenth? And what evidence could such a comparison of complexity possibly be based on? Instead of providing »historical reasons,« we may well focus, as Foucault already did in Les mots et les choses, on some symptoms for an intensifying epistemological tension as it came up during the eighteenth century. What got lost soon after 1700 was the intellectual trust and the existential premise according to which the human mind in general and each individual mind in particular were capable of producing fully adequate (and therefore consensual) representations of the world and its objects. Denis Diderot’s and d’Alembert’s heroic project of bringing together, in a systematizing structure, all available knowledge through their Encyclopédie, ou Dictionnaire raisonné des sciences, des arts et des métiers, may have been the one institutional and intellectual place, almost paradoxically, that made this tension particularly visible. For it relied on the double-leveled condition that, firstly, the use of unprejudiced human reason was enough to produce adequate representations of all phenomena and that, secondly, the accumulation of such representations would end up revealing a general and permanent grid of all knowledge (an epistemological grid that would mirror the ontological grid of the cosmos and its objects).

The editorial reality of this vast enterprise, however, turned out to be much more centrifugal. Not only did Diderot and d’Alembert soon understand that it was inevitable, by the advanced intellectual standards of their time, to let different authors describe seemingly identical phenomena from different perspectives; in reading manifold entries from the Encyclopédie we also sense an obsession with »ontological dispersion,« that is with a fear that phenomena thought to belong to coherent fields of knowledge and practice would no longer obey such presuppositions of order. I have always felt that there is a representational strategy (or perhaps just a preconscious inclination) in the »planches,« that is in those volumes of the Encyclopédie that supplemented its textual entries and descriptions with detailed etchings of multiple objects, to bring centrifugal objects together into the (of course imaginary) spatial order of frames, buildings, and workshops. The spaces of the bakery or of the butcher’s shop, the study of the writer or the office of the merchant seem to suggest and, for a brief historical moment, even to rescue an impression of unity that was then vanishing in the cognitive realm.

We can detect a similar dynamic in the history of the concept of »style« and in the practices that it informs. There is a famous sentence by Georges-Louis Leclerc, Comte de Buffon, published in the Encyclopédie—»le style est l’homme même«—which, over decades and centuries, has become the lemma for a conception of style as a range of multiple forms expressing a multiplicity of individual dispositions. By contrast, it had been Buffon’s original intention to counter and contain precisely this profusion of individual styles, by claiming that there had to be one, and only one, correct human way of seeing and arranging things in a logical fashion, one and only one logical order that would guarantee an adequate perception and representation of the world. »Style« for Buffon was a normative concept set against individualizing perspectivism.

It was typical of an eighteenth-century frame of mind, different again yet equally powerful, that tackled the same epistemological problem, i.e. it was typical of enlightened Materialism, that philosophes like Julien Offray de La Mettrie, Claude Adrien Helvétius, and Diderot were obsessed with the possibility of proving that the human senses were well capable of producing truthful renditions of a reality lying outside the human mind and the human body. This of course was the reason for their—and for so many of their contemporaries’—fascination with the blind and the deaf: it came from their will to show (or rather, to imagine) that even the loss of an entire perceptual dimension would not fully sever the human mind from the objects lying outside itself. A philosophical—and thus either overly friendly or definitively castrating—reading of the novels written by the Marquis de Sade could arrive at the converging impression (as Foucault did), that is at the impression that they were complex intellectual experiments trying to confirm the cognitive reliability—and dignity—of the human senses.

Finally and above all, Immanuel Kant’s critical work was primarily motivated by the urge to show that, ultimately, the human senses, the human mind, and human judgment were well qualified to experience the objects of the world in an adequate way—and to deal with them on such a basis. This project, however, only accounts for half of Kant’s unique importance within the history of Western philosophy. If he was, on the one hand, the last truly great philosopher who could persuade himself that humans had the necessary equipment to capture the world of objects; nobody on the other side of the epistemological divide pushed possible doubts against such optimism that early and that far. But I will still refrain from saying that similar cognitive and epistemological tensions were the historical “reason” for the emergence of second order observation as a habitual feature of intellectuals’ minds in the late eighteenth century—although one can of course plausibly claim the existence of such a relationship. It is also obvious that similar forms of skepticism and similar reactions to it may have articulated themselves in different national and regional cultures at different chronological moments. In this sense, too, what I am trying to establish here has the status of a potential frame narrative.

Read the full text and more about this topic in the publication Design of the In/Human.