Robert C. Priddy

Public lecture for the Norwegian Magistergrad ('Klassisk Magistergrad' equivalent PhD) - July 1968. Institute for of Philosophy, University of Oslo

In order to speak about the nature of 'observation' and 'understanding' in respect of the social sciences, I have chosen the approach of considering the traditional context of thought in which the words have mostly been, and frequently still are being, used. It is not so much -or not primarily- in everyday language that the concepts these expressions denote are problematical, but in the language and praxis of the sciences to which they variously belong, or from which they have arisen.

I shall begin by sketching a prevalent view or notion of observation that derives from the practice of science - particularly the natural sciences. This view has, from time to time, been defended by scientists engaged in talking about the ways in which their knowledge is derived, as well as by certain philosophers of science -particularly thinkers in the empirical and positivistic traditions. The notion of simple, descriptive observation to which I refer has been developed and expressed in a great number of ways. For example, it is dealt with in a simple commonsensical way by G.E. Moore and, conversely - in a most sophisticated and complex fashion by certain of the logical positivists. I shall concentrate mainly upon the following chief themes:-

Firstly, that simple and direct observations can be made and expressed in language so that nothing but the observational content is included.

Secondly, - being a variation of the first - that observational statements have a purely descriptive content, being simple.

Thirdly, being an extension of the first and second themes, such statements-being only descriptive are not normative. Thus they contain no moral exhortations, but also, they are supposedly free of any evaluative judgement or of so-called subjective content. Anyone who understands the language, in which they are expressed could, in principle, verify them.

Fourth, being a rider to the proceeding themes, that such statements are required to have been made by a supposedly 'disinterested' spectator to a particular state of affairs, who - in the function of observer - simply operates as a neutral registrar, that is, one who simply registers what is the case or simply records a matter of fact.

Finally that such simple, descriptive observations are pre-explanatory and that they have neither any hypothetical nor theoretical content, nor do they thereby rely upon any form of pre-judgement or subjective interpretation.

The product of such observations are in general spoken of as data or givens. Before attempting a critical discussion of this general view of simple descriptive observation, I find that some mention can profitably be made of the interests in which such a view has come to be held. I shall briefly consider the history of the notion of observation itself.

At the height of Grecian culture, observation is already fairly widely employed as a means to knowledge such as in the medicine of Hippocrates, the history and ethnology of Herodotus and, above all, in the biology and other natural scientific pursuits of Aristotle. Yet observation was by no means regarded as a master principle, but the tendency was rather still to regard descriptions of the transitory or sensible world as opinion or doxa, and seek instead to base knowledge upon intellectual intuition of a supposedly unchanging and ordered universe. This is the Grecian view of pure theoretical knowledge or epistime, whereby the intellect grasps the principles of the cosmos in logos or logical intuition, principles which were not at all evident to common sense observations.

With the development of modern natural science, however, thinkers tended more and more to emphasise the importance of observations or empirical data. From the time of Francis Bacon and onwards, the various thinkers who tried to explicate the methods of the new sciences came to regard the practice of observation, the reference to empirical data, as the sine qua non of knowledge.

Without a fundament in observations there could be little more than misleading speculations. Knowledge now came widely to be thought of as scientific knowledge, the model being natural scientific knowledge. The laws to which such sciences attained were thought of as being inductively generated from observations, variously aided by experience or, as Hume might have put it, by natural expectations of finding certain events following others. It is by comparison of a number of simple observations and, finally, by the observational verification of hypotheses suggested by original observations, that empirical laws are generated.

Empirical laws were statements as to which events would follow other events under given conditions. If a prognosis proved to be correct upon final observation, one was considered to have obtained knowledge of natural and more or less universal laws as to how nature changes. Such a philosophy of science finds a good deal of support in the views of Hume, with his demand that all ideas should be reducible to simple impressions – which is to say that they must be traceable back to direct perceptional observations.

Now, as the German philosopher Jϋrgen Habermas has pointed out, modern science also took over the Grecian view of the desirability of the disinterested, theoretical attitude and its related assumption of a structured world in itself. This notion of an objective world, behind the scenes, has frequently been called objectivism. Theory was a map or copy of the unchanging structure of the logos. Modern scientific knowledge, thereby, was thought of as a body of observations united in theory, from which all changeable factors or variables are purged. Observation is, by now, the cornerstone upon which theoretical structures are erected.

But further than this, modern science came to assert a separation of fact and value, of is and ought, such that its theory should be purged of all evaluative or culturally-variable judgements. This view, we may also remark, was foreign to the Greeks, for whom theory was precisely the means by which fact and value, the given structure of cosmos and the human strivings to come into harmony with it, we're unified.

The chief point I want to make here is that most current notions of observation are inalienably connected to a certain type- of view on the methods of natural science and, above all, the aim of these methods -that is, the goal of science as universally valid knowledge of the unchanging structures in nature, or of the principles according to which structures necessarily change. Observation here does not itself constitute understanding of phenomena observed. According to this view, understanding first comes about when knowledge of constants or permanent structures of the natural world is obtained- when a verifiable theory has been constructed upon the basis of discrete observations. Or, in a more sophisticated form, when principles are described according to which such structures are governed. In theory, the occurrence of transitory phenomena and the connections or relations pertaining between them must, in order to be understood, be related to or subsumed under constant governing laws... or generally shown to be variations upon and within some fixed fundamental structure.

Such a notion of observation and theory were already carried over into the anthropological sciences by the European initiator of sociology, Auguste Comte. Comte proclaimed positivism as the philosophy of the social sciences, which was much the same as regarding society to be a natural entity, governed by universal natural laws. The notion which I regard as crucial and fateful here, one which underlies a large part of sociology even today, is that the aim of sociology is to reveal an underlying structure in society whereby human actions and social changes may be explained or accounted for by reference to such human interactions and the social institutions in which these result are thereby thought of as being understood only when they can be seen as examples of or cases of universal laws operating within a fixed basic structure - or laws known variously as ‘empirical generalisations’ or even ‘natural principles’. This view, with all its various riders constitutes the essence of sociological objectivism. Observations are thus but simple descriptions upon which theories can be built, so as to provide a genuine understanding of society. Comte regarded the natural scientific method and its theoretical ideals as eminently suitable to the study of human interactions and society. And while Comte’s positivism suffered at least a nominal defeat, his objectivism has persisted in general sociology up to the present day.... in the views of Pareto, Durkheim and many subsequent thinkers. To give a contemporary local example, in the notion of a pure social science as defended by Johan Galtung.

Seeing how observations are so closely geared to that of pure theory, it becomes the clearer why scientific empiricists in the naturalistic, and objectivistic or positivistic traditions are so insistent upon the purity, the simplicity, and the infallible directness of observations. For, without such a pure and simple basis - free of subjective interpretations and evaluations - the certainty or infallibility of the whole superstructure of theory is cast into doubt. Put in another way, if one fails to secure acceptance for the general principle that basic, descriptive observations can be made which are universally valid in the sense Of being Allgemeingǖltig, the entire way of making reality intelligible or of understanding reality as represented by empirico-analytical evidence science becomes questionable.

The philosophy of neo-positivism, which arose in the earlier part of this century, therefore tried to secure the primacy of the so-called principle of verification, which, in general, demands that, if statements are to be regarded as meaningful, they must be reducible to observations (‘reductionism’). They tried to set up observation and the concurrent idea of pure theoretical science - which they spoke of as Unified Science- as the only correct way of making reality intelligible - regardless of whether it is physical or social reality that is to be understood. I think there are good grounds for saying that the neo-positivists succeeded mainly in showing that, once a notion of pure descriptive observation is adopted - once it is accorded such a cardinal and crucial role as the fundament of theory and science altogether - many virtually insolvable problems arise. These include the difficulties of showing that discrete or simple atomic observations are possible and that they are not subject to individual subjective variation, and also of explaining how observational checks or re-observations can be made where the phenomena observed are subject to change, of making individual sense observations publicly accessible and verifiable, and the consequent problems of solipsism.

A commonly-adopted solution to the philosophical dilemmas attendant upon any strict adherence to the notion of staple and pure descriptive observations is to recognise that observations and the statements in which they are expressed can be only relatively simple or descriptive, only relatively discrete and relatively Allgemeingϋultig. This involves the social scientist in rejecting the notion of pure theory, the notion that there is only one correct way of making society and social forms of life intelligible.

The notion of simple, descriptive observation is closely bound up with the philosophical attitude of objectivism. In sociology, this thesis took or various forms. John Stuart Mill, the English philosopher of science who developed the notion of simple descriptive observation and the methods of comparison by agreement, difference, residues and concomitant variation, maintained- for example- that there is no fundamental difference between the principles according to which we explain- and that is to say, properly understand- natural changes and social changes.

Somewhat similarly, Emile Durkheim is an objectivist in that he finds extremely fruitful the idea that social life should be explained (or- understood not by the notions of those who participate in it but by more profound causes which are unperceived by consciousness.

Again, fundamental to the thought of Vilfredo Pareto is the view that the ideas of participants must be discounted as more likely than not to be misguided and confusing. Instead Pareto insisted that the sociologist develop his own concepts, independent of those current in society, but having a strictly empirical reference. We see particularly clearly in the work of Pareto how misleading an idea of simple descriptive observation can be. He considers that we can find certain constant ideas in different societies, which concepts he calls residues - or what is left when the subjective variations are left out of account. By simple observation one notices, for example, that in several societies water is used in baptism and other rites for giving the individual a sense of being morally cleansed. That a Christian says baptism removes Original Sin and – in certain other religions the priest gives other grounds for the cleansing power of water - is regarded by Pareto as inessential. The observable constant is that water is used for moral cleansing.

But, Peter Winch makes a clear commentary upon this view:-
"....ideas cannot be torn out of their context in that way; the relation between idea and context is an internal one. The idea gets its sense from the role it plays in the system. It is nonsensical to take several systems of ideas, find an element in each which can be expressed in the same verbal form, and then claim to have discovered an idea common to all systems."

In all these cases, understanding proper supposedly first arises only when a theoretical explanation of the occurrence and purposes of human actions is given... when they are interpreted in terms of objective categories and theoretical generalisations, themselves developed by comparing and categorising observations from various similar areas of social life. The terms by which human actions are generally understood, the motives and purposes of acting individuals and the ways in which social phenomena are understood through our everyday intercourse with them are to be rejected. Social phenomena are not to be grasped from the point of view of participating social subjects but from an objective theoretical viewpoint. According to objectivism, therefore, understanding of social phenomena is not the same as mere observation of them, understanding proper arises only when one Ins gone beyond the stage of discrete observations. But, since this supposedly objective understanding rests, in the first and last instance upon observations, one is lead to insist upon the notion of sheer, descriptive observation of discrete states of affairs by disinterested spectators who maintain the theoretical attitude. If not, the objectivity and neutrality of the theoretical superstructure is - by the same measure - cast in doubt.
The notion of observation I’m considering, therefore, does not so much arise from any adequate conceptual analysis o£ the praxis of observation - how we actually do and can observe social phenomena - but from the requirement that neutral and discrete, simple observations be possible. This demand follows as a consequence of adopting an objectivistic theoretical schema of scientific explanation in the first place. In this, however, one has conveniently overlooked the fact that the social scientific observer is himself a social being, partaking in society. For it is by virtue of his prior knowledge of social ways of life and what we call normal or ordinary understanding of actions and social phenomena altogether, that he can at all make observations, or even express them in language. As Wittgenstein has indicated, the ability to use language is inextricable from the understanding of a way of life or Lebensform.

Moreover, scientific observations are not made haphazardly, one does not go around observing aimlessly, but one does so guided by the theoretical interest of developing fruitful hypotheses, and eventually theories, about particular social phenomena. This interest is itself a derivation of more fundamental interests, whether technical, practical or political. This is to say that observations - data or givens- are also equally capta or ‘takens’. Observation is selective. Prior to the formation of any theory, there must be observations of some sort - and such observational selection of data is guided only by common experience. It is on the basis of each individual observer's ordinary social comprehension - which he has through himself being a social member - that observations are made and data selected. This is often alluded to in books on scientific methodology - without their giving any further details, by saying that the scientific observer - prior to developing hypotheses - is guided by experience and common sense, or by expectations an intelligent person develops from repeated dealings with the subject at hand. One cannot develop fruitful hypotheses - to whatever purposes - upon just any observations.

To draw a parallel to Kant: while the arguments of the ‘transcendental logic’ showed how experience or empirical observations in the natural science rest upon prior categories of the understanding, what is required in the case of the social sciences is to recognise that observational judgements about social phenomena are founded upon everyday categories of social understanding, which often vary considerably with the culture, group or individual involved.

The influential sociologist Max Weber appreciated much of this difficulty, not least since he was conversant both with the theory of natural science and with the philosophy of historicism, such as propounded by Dilthey and Rickert. In the German tradition since Kant, knowledge fell into two categories - radically different and apparently unbridgeable. On the one hand there was knowledge of the world of nature, or of what Kant called the phenomenal world, where one tended to view nature as a closed system which must be treated only empirically. On the other hand, historical knowledge - including knowledge of society - held to the Kantian world of Geist or spirit.... hence the term Geisteswissenschaften. This latter sphere was accessible primarily by understanding or Verstehen. The thesis of Verstehen was that the world must be directly or intuitively understood from the point of view of the participants, from the so-called subjective point of view. But, since humans in different historical epochs or in different cultures made reality intelligible in vastly different terms and through various ideas, it followed that generalised theoretical categories were unusable in reaching proofs of causal relationships in the human, social and cultural fields.

Weber attempted to bridge the gap between these two trends in the region of sociology. His main approach was to assert that generalised theoretical categories are perfectly admissable provided one does not thereby believe that they will enable one to attain to a complete picture of the ontological reality or being of any phenomenon.

Such categories must meet the logical requirements of the theoretical procedures common to natural science, but also they must embody the subjective categories of understanding peculiar to human cultural society.

Observations of social actions must be based upon actually-existing meaning in specific social instances where an actor intends something. This is grasped by the observer in immediate interpretative understanding. When we see somebody bringing an axe down upon a log and splitting it into pieces we directly understand, in view of the specific situation, that what he is doing is chopping firewood. There is also a more complex type of understanding where similarity we grasp a person's intentions within a more inclusive context of meaning. For example, where we understand that the person before us is chopping firewood in order to have dry wood available in the winter months, or where he is chopping it for a wage for an employer.

Such meaning is - for Weber- the chief content of basic observations in social science. For Weber, then, understanding and observation are so far "one and the same". Weber asserts that in all cases, understanding involves the interpreta1 grasp of subjectively-intended meanings. But no account of social action based on the grasp of such meaning alone can be considered to give an adequate sociological understanding. What is also required is a causally-valid explanation of the actions observed and grasped in understanding. I am suggesting here that Weber, having attempted to avoid the pitfalls of objectivism, simply falls back into them in a more sophisticated way. For causally-valid explanations are those which consist in subjective interpretations of action that have been empirically tested and verified as to their correctness by comparison with similar types of action in similar situations. They are judged causally-valid explanations if it can be shown that similar actions actually occur in similar situations.

In all this one again relies on the objectivistic notion that actions - in whatever society or epoch - are governed by constant laws. In Weber’s view, when one can successfully predict an action’s outcome by referring to a model developed from observations in another but similar social situation, one has attained to such a law and thereby reached an objectively-valid understanding of the social phenomena involved.

The English philosopher, Peter Winch, has shown – in his book The Idea of a Social Science – how Weber's demand, that we must supplement our so-called subjective understanding of social action with another type of understanding - or by empirical proofs that we understand correctly in the first analyses - is misguided and superfluous.

Winch refers to Wittgenstein's example whereby we are asked to imagine a society in which wood was sold in the following way. "They piled the timber into heaps of arbitrary, varying height and then sold it at a price proportionate to the area covered by the piles. (And what if they even justified this with the words: '0f course, if you buy more timber, you must pay more'?) The important question for us, says Winch, is in what circumstances could one say that one had understood this sort of behaviour? Weber often speaks as if the ultimate test were our ability to formulate statistical laws which would enable us to predict with fair accuracy what people would be likely to do in given circumstances. But, Winch continues, with Wittgenstein's example we might well be able to make predictions of great accuracy in this way and still not be able to claim any real understanding of what those people were doing.

Instead, therefore, of looking to statistics or comparative averages of behaviour which apparently conforms to the same type in order thereby to make one's understanding valid as the causal explanation of an act, one needs instead — wherever difficulties of explanation arise — nothing more or less than a better interpretation of the phenomenon. The whole question of understanding is a conceptual one, not an empirical one. The probability of occurrences of certain acts tells nothing new about their meaning, no more than an empirical knowledge of the frequency of vowels in a language will help one in understanding the meanings conveyed through it.

My way of treating the difference between observation and understanding: in sociology has mainly been to sketch the genesis of the terms and the notions so often inseparable from them so as to indicate in what interest the distinction arises. This, of course, by no means exhausts the subject. But one preliminary conclusion can, I think, be made: that the distinction between simple descriptive and value-free observation on the one hand, and the participant understanding of a way of life and the concepts germane to it on the other hand, is maintained so as to preserve the fundament of objectivistic theories of society.

Yet in fact - as Weber largely appreciated - the distinction between observation and understanding is artificial when applied to social action and social institutions. In the praxis of social scientific observation, as I have briefly tried to show, it is impossible to arrive at a purely descriptive observation, any sheer pre-explanatory registering of data. As I suggested when speaking of logical positivism, observations can only be regarded as being relatively free of subjective interpretations in the sense of being more or less Allgemeingϋltig.

What now of the notion that observations must be formulated in simple and discrete statements corresponding to simple and discrete states of affairs? This depends upon the correctness of the view that social acts and phenomena can be understood in isolation, can be intelligibly described independently of each other and without reference to other phenomena. Once again, this involves the concept that social phenomena co-exist in such a way that the actual relations between them are external, the one phenomenon can be viewed independently and be what it is regardless of the existence of other phenomena. Contrary to this, however, I would maintain that -at least as far as social phenomena are concerned, this is never factually the case. To give a simple example in support of this: how can one understand the social phenomenon 'a school' without also understanding the internally-related notions of what a 'teacher' is, what it is to be a 'pupil' and what it is to learn and be taught? How can one make observations about the number of successful candidates at the yearly school examination without comprehending what it is to sit an exam, what it is to correct examination papers and what intellectual criteria are employed in judging that a candidate has passed? Observations, therefore, can only be spoken of as being relatively simple and discrete statements about certain aspects of a more or less integrated way of life or Lebensform; they represent moments in the social scientific observer's understanding of that way of life.

Finally, some mention of the problem of value-freedom. Observational statements may, of course, be non-normative – i.e. in that they do not consist in prescriptions of how one ought to act or how the observed phenomenon ought to be. But if one accept the view that all observations necessarily take place within a lived understanding of a social way of life, it is clear that they cannot be completely devoid of normative contents, for the very concept of a norm contains the notion of doing things, interpreting acts and events in a regular and common way. To claim that observations or understanding must be completely devoid of all normative elements whatsoever is the same as making them devoid of all value - and even meaning - whatsoever: a phrase which speaks for itself.


copyright Robert C. Priddy, 1968.