Science and the social construction of reality

1. Introduction

In their famous work The Social Construction of Reality1, Peter L. Berger and Thomas Luckmann analyse ‘knowledge’ as a social construction. In doing this, they obviously do not mean knowledge in the fullest sense, but rather whatever is socially constructed as knowledge and reality2. This, however, includes everything. There is no neutral kind of knowledge — all is socially constructed. What, then, is the status of such things as science that are believed to reach more objective knowledge? What would be the status of a hypothetical system of knowledge that was in fact completely true and comprised true knowledge? Could such a thing be attained?

In this essay, I intend to make a survey of some aspects of Berger’s and Luckmann’s theory and proceed to make my own observations. In particular, I will look at how their model of social systems of knowledge can apply to all kinds of knowledge — including the theory itself — and what effect this has on the credibility of the theory and to that of all theories that seek to obtain objective knowledge.

2. The basis of institutionalisation

We will here mainly be concerned with social systems of knowledge that already exist or which are created in other ways, but for contrast if nothing else it will be useful to look at how the social construction of reality normally happens according to Berger and Luckmann’s model. According to them, institutionalisation begins from habitualisation. When one person interacts with a particular environment, they will develop routines that save them effort by sparing them having to choose every action anew in repeating situations3. Institutionalisation, in turn, is when such routines form an interacting system where particular kinds of routines are associated with particular roles people can act in4. Such a structure enables the participants to save effort not only in not having to decide each action anew every time, but also in being able to anticipate what others will do5.

Institutions in a society gain what Berger and Luckmann call objectivity, because when they are transmitted to the next generation, they at first appear as simply reality. They become something that is accepted automatically as opposed to something (such as one’s solitary routines or rudimentary institutionalisation born from scratch between people of the same generation) that has a known origin and can be changed. This objectivity also appears in a way analogous to that of the objectivity of nature; the social institutions comprise a world that exists independently of the individual and cannot be changed by them at will, and which has existed before they were born and will probably go on existing after they die.6 However, this does not in itself ensure that members of society adopt its institutions. The very fact that the origins of the institutions are not immediately present to a generation that did not originate them also makes their reasons for existing that much distant. This is why legitimation is needed.7

3. Legitimation and universe-maintenance

Legitimation of the social order is what helps it seem to make sense, both on the level of the totality formed by the institutional processes from point of view of the individual taking part in them, and on the level of an individual’s life as it passes through the institutionally defined phases. As such, it produces new meanings that integrate those already attached to the institutional processes. Legitimation gives both an element of values and of knowledge, not only what must be done but also an explanation of what the world is like to cause this to be so.8

There are, theoretically speaking9, four distinct levels of legitimation for institutions. The most automatic level, incipient legitimation, is pre-theoretical and based on the meanings encoded in the language and its vocabulary itself. The second level is theoretical in a rudimentary sense, containing pragmatic explanatory schemes closely related to action, expressed in such forms as sayings, moral maxims and stories. The third level is something we will be concerned with later in this paper, because it is the explicitly theoretical level on which science will have to be placed (although Berger and Luckmann do not clearly make this observation about science at this point in their text). This level tends to contain such advanced knowledge that it is left for experts to handle rather than for everyone to understand. The fourth level is that of the symbolic universe, which integrates all human experience in a single frame of reference. This frame of reference is symbolic in the sense of referring to something outside everyday experience, and a universe — universal — in the sense of containing all human experience, with nothing left outside its scope.10

The talk of symbolic universes brings us to the idea of universe-maintenance. Like institutions, symbolic universes need legitimation in certain circumstances that might bring them to question. This would never happen if the symbolic universe could always be taken entirely for granted (as it is in a stable ideal situation), but this is in practice impossible, and there will always be problematic situations. One cause for this is that transmitting the universe to the next generation is never perfect, and people will end up with slightly different views of the universe in spite of society’s best efforts. To this contributes the fact that, as mentioned above, the symbolic universe refers to something outside of experience, which it is naturally not as easy to present to the next generation as concrete things are. Another problem (which also adds to that of the socialisation of the next generation) is that of groups within society adopting and to themselves objectifying deviant views of reality, which thus become theoretical and practical rivals for the mainstream symbolic universe. Historically, it has in fact often been the case that the upholders of symbolic universes have been motivated to systematise and theoretically conceptualise their systems of belief only when faced with such challenges — Berger and Luckmann give Christological controversies/heresies in the early church as examples. Another similar situation is, naturally, when the society comes into contact with another that maintains a different symbolic universe from its own (as any truly separate society will, of course), showing on a large scale the possibility of a quite different universe, and by making do without them demonstrating in practice the arbitrary and not so inevitable nature of some of the things taken for granted by the first society. If there is such a good reason why we must do things this way, how come they can do otherwise and still get by just fine?11

Berger and Luckmann speak of conceptual machineries of universe maintenance, systems of abstract reasoning developed to explain the superiority of one’s own system. These represent a systematisation of legitimations that were already present in society in simpler form. Despite their ‘logical’ character, they are still not applied in a purely intellectual manner. Rather, whose explanations win depends more likely on matters such as military might; it is not really that the theoretical constructions win, but rather that the winner gets to spread their version.12 On my part I could add the observation that this is hardly surprising, considering how hard it is to by pure reasoning convince of anything persons who are determined to hold a particular opinion, possibly because of ulterior motives. As far as more specific kind of conceptual machineries of universe maintenance go, Berger and Luckmann mention as examples mythology, theology, philosophy and science13. In a different classification, they mention therapy and nihilation. In a conceptual sense, the former means a system that explains away deviance of individuals as the aberration it should be seen as and also entails a theory of an appropriate curative process, as well as preferably means of pre-emptive detection and preventative measures. Nihilation, in turn, applies to ‘threats’ from outside the society, which cannot be assimilated and so need to be explained in a way that makes them no more a threat.14 We will return to nihilation in the next section.

Now that we have examined some of the basics of Berger and Luckmann’s theory on a general level, we can examine its applications to the question of scientific knowledge. In the next section, we will start off by seeing what the authors themselves mention about science in relation to their theory, and proceed by examining what can be said about their theory itself in its own light.

4. Scientific knowledge as socially constructed and maintained

Berger and Luckmann do not have much specifically to say about science or specific scientific disciplines, but some of the things they do say seem rather unflattering. One example is given when they discuss the maintenance of bodies of knowledge that become autonomous sub-universes of meaning. As these become more numerous and distinct from each other, it becomes necessary to draw a line between the insiders who are initiated into the particular universe and the outsiders who are not; the outsiders must be kept out and the insiders must be kept in.15 Berger and Luckmann give an example from the field of medicine.

It is not enough to set up an esoteric sub-universe of medicine. The lay public must be convinced that this is right and beneficial, and the medical fraternity must be held to the standards of the sub-universe. Thus the general public is intimidated by images of the physical doom that follows ‘going against doctor’s advice’; it is persuaded not to do so by the pragmatic benefits of compliance, and by its own horror of illness and death. To underline its own authority, the medical profession shrouds itself in age-old symbols of power and mystery, from outlandish costume to incomprehensible language, all of which, of course, are legitimated to the public and itself in pragmatic terms. Meanwhile the fully accredited inhabitants of the medical world are kept from ‘quackery’ (that is, from stepping outside the medical sub-universe in thought or action) not only by the powerful external controls available to the profession, but by a whole body of professional knowledge that offers them ‘scientific proof’ of the folly and even wickedness of such deviance. In other words, an entire legitimating machinery is at work so that laymen will remain laymen, and doctors doctors, and, if at all possible, both will do so happily.16

This paints a rather unflattering picture; indeed, it could perfectly well describe the workings of some society we would see as primitive while condemning its medicine as hocus-pocus. Yet there is no particular basis for denying that it fits our current society and medical profession. This gives a rather dim view of things, as the reasons why the status of medicine holds such an esteemed and authoritative position apparently have nothing to do with its scientific and supposedly reliable basis. Instead, they are based on the same kind of mechanisms of legitimation as any arbitrary system of beliefs that would occupy the same position in society could have. Indeed, other such systems would also have their own ‘science’ backing them up. It is easy enough to see that the same type of problem extends everywhere in science. Science cannot claim to exist without a social system to support its ideas, it cannot maintain its status without mechanisms of universe-maintenance, and therefore all of Berger’s and Luckmann’s arguments apply to it as well as to anything else. Every field of science has its own specialists who are specifically initiated into it, and they will be called out for going outside the norms, while laymen are not allowed to speak with the same authority on the matters the science is concerned with. Further, the fields of science have elaborate theoretical justifications for their belief systems, designed to justify their position above their theoretical rivals. This is necessarily so because a field of science always claims to be the superior system in explaining what belongs to it. The same applies to individual theories, which is where we get to Berger and Luckmann’s own theory. It is evident with a little reflection, and intriguing, that when they speak of theorising and legitimation, they are doing those very same things at the same time, and justifying their own system. Social sciences are hardly exempt from the observations made above. The most interestingly meta part of Berger and Luckmann’s description of their theory comes when they speak of nihilation as a means of universe-maintenance. As sketched above, nihilation means explaining away the views of other groups into irrelevance — as the name says, making them be nothing at all. The members of the group with the rival views are assigned an inferior ontological status, such as that they are barbarians or “not human”, which also enables not taking their views seriously. Secondly, the rival views are actually explained in terms of the group’s own beliefs, incorporated into its own universe, translated into its language. This way, it comes to be the case that whatever the others may say to deny the symbolic universe really just affirms it.17

What makes this bit of theorising so interesting is that it is doing the exact same thing that it describes. Much of the rest of this book is dedicated to this too. Simply put, in Berger and Luckmann’s model no symbolic universe is an objective reality, but they all claim to be — and what Berger and Luckmann write here acts as a rebuttal of those claims and nihilation of those symbolic universes. In their theoretical explanation, we are talking merely about man-made, symbolic universes, believed in by naïve people who lack the theoretical understanding they possess. Properly understood, within the framework given in this book, they become merely symbolic universes, not the objective truth they claim to be, and are thus nihilated. The particular explanation of nihilation explains how the symbolic universes can indeed seem to be better than their rivals and cover everything, when it is in their nature that they can perfectly well make false claims. Rather than reacting to any particular claims that would threaten their view, Berger and Luckmann create a model that already contains its “rivals” and explains them away before anything actually rises to challenge it. Of course, it is the nature of their theory as they intended it to do just that.

What can we conclude from this? One unwise possibility would be some sort of relativism. Because all ‘knowledge’ true or false is socially constructed and legitimated with the same kinds of mechanisms, one could conclude that it is all equally true. This is, of course, nonsense. Even this explanation would have to consider the idea that all knowledge is relative as more valid than the rival ideas according to which it is not, lest it affirm and deny the same thing at the same time. In general we also know that, much as society can affect how people see the world, there is still an independent world that we do see this way, to which the very fact that systems of legitimation are needed is partly due. Even Berger’s and Luckmann’s theory is entirely dependent on a world that is objective beyond human beings’ seeing it that way.

5. The quest for objectivity

Now that we have established scientific theories including Berger’s and Luckmann’s own as being the same sort of constructed universes as all other ‘knowledge’, it is time to look at what exactly this means when they are also seen as aiming at objective truth. The main point I will be making is that whether knowledge is a matter of social construction and whether it is reliable knowledge have no necessary bearing on each other in either direction — but it is nevertheless important to understand the socially constructed nature of the given sub-universe of meaning, even if it is the one of a field of science.

In truth, phenomena related to the notion of the social construction of reality have always been a major issue in attempts at achieving objective knowledge. Today’s science condemns many past mainstream ‘scientific’ notions as basically nothing more than legitimations of rationally baseless world views. An example would be the justification of racism and exploitation based on ‘biological’ notions of race. One could ask whether current views on race are not similarly affected by the desire to see everyone as equal, which could actually be seen as a good thing. It is in fact likely current ideas are also having an effect on scientific knowledge, but this does not have to make them purely pragmatic constructions. Further on the topic on the problems of achieving objective knowledge, even more of these have been on a deeper level that is not so much what Berger and Luckmann speak of as it is what is behind many of those things. I am referring here to the human tendency to use the data they gain from the outside world to confirm their notions more easily than to refute them, and to the limits of a person’s perspective in general. In other words, it is always very hard to test one’s hypotheses reliably against reality without simply falling to the trap of seeing what you expect or want to see. This is a key factor behind why the socially constructed reality can exist at all, as it requires seeing things in accordance with a certain preconceived model. For an individual, however, such a model need not be socially enforced. It can just as well be born out of heir own mind, which may even happen as they are observing the phenomena. Of course, while the presence of this tendency may make the practice of science and in general obtaining information about non-immediate facts difficult, the absence of it might just make those things impossible. It is our way of making sense of the world, after all.

Whatever reliability science can claim is based on its awareness of and ability to control the above tendency. This, if anything, is the distinguishing mark of ‘real’ science. Controlled studies, replication of results, peer review and all the methods of the individual fields of science are designed to keep people from being able to assert results that are not backed up by reality. Even so, the process of actually finding out anything is complicated and arduous. Established notions and theories get overturned — but this happens based on new evidence. Biblical creationism was not disproved because it went out of fashion in science. It went out of fashion because it was disproved, because geologists developing a greater understanding of the Earth’s surface found the planet could not possibly be only thousands of years old, and because evolution was brought in as a rival theory of the origin of species and later proved to be the one that fit the evidence. This was met with resistance because the old view was so embedded in social reality, but the new view prevailed because scientists had begun to really look at what they were supposed to be studying, and could not ignore the evidence. Of course, what they were doing instead of looking at things through the old world view was to look at them through another, different one. Nevertheless, when they saw this new view fit the world better than the old, they did not simply conclude this because they were already motivated to do so.

Every system of legitimation aims to be scientific like this. They all want to show that things are objectively the way they are believed to be. However, no such system can be scientific as long as it only aims to show what is already believed to be true. This almost invariably leads to its begging the question and simply taking any arbitrary path it can find to its desired destination. This is why actual science sets out experiments to falsify its hypotheses, not verify them. Uncountable generations of people seeing what they wanted to see has taught it this lesson. Even this testing cannot alter the fact that scientists still have to look at the world through certain theories. But science tests out theory after theory, and keeps testing each one, only keeping it as long as it fits better than any alternatives. (If someone would say this is not how it really works because of something in the sociology of science, I know those kind of objections and also counter-objections to them well enough to just think I don’t need to go into that here. As I see it, perspectives like Kuhn’s add complications along the way but change little in the end.) Well-entrenched theories are of course hard to get rid of, but enough contrary evidence will eventually topple even them. Berger and Luckmann do not really discuss this, but they hint at it in their discussion of sub-universes of meaning and their associations with social groups:

Especially on the theoretical level it is quite possible for knowledge to attain a great deal of detachment from the social and biographical interests of the knower. Thus there may be tangible social reasons why Jews have become preoccupied with certain scientific enterprises, but it is impossible to predict scientific positions in terms of their being held by Jews or non-Jews. In other words, the scientific universe of meaning is capable of attaining a great deal of autonomy against its own social base. Theoretically, though in practice there will be great variations, this holds with any body of knowledge, even with cognitive perspectives on society.18

It may be exaggerated to say that it is always impossible to predict scientific positions based on someone’s social position, but the point holds nevertheless: science has its own, quite genuine rules, and those are rules for actually finding out the truth.

We can now take a second look at the example of the medical profession quoted in section 4. Let us first assume that modern medical science is ideally scientific and its practical applications are direct, rational results of that scientifically obtained knowledge. With this assumption, we find that it has every right to legitimate itself in the ways described. If it is scientifically based, it represents the best knowledge available and the best chance of good results in looking after people’s health. Only experts can competently administer it, so it is just as well the distinction is made between them and laymen and they are given authority on such matters. As for alternative views, anything that is actually proven to work will be accepted as part of medical science, meaning that anything that is left outside it has either not been tested or has been found to be ineffective. It is known that without controlled experiments, just about any kind of treatment can appear to be effective because of confirmation bias and the placebo effect. On the other hand, it is also known that there are many things that cannot be cured by the placebo effect but need real treatment. Additionally, it is ethically questionable to make a business of selling people ineffective treatments as genuine. Therefore, the practice of limiting these ‘heresies’ is perfectly correct. It is not merely that there exists a theory that explains why mainstream medicine is reliable and alternative medicine is not. This theory is also quite valid. Now, if we drop the pretense that modern-day medicine is a perfectly rational and disinterested discipline, we will find it has its own social and economic sides that interfere with the idea of pure science being used solely to help people. But we will still find that sound science has a place in mainstream medicine, unlike any of its alternatives that I am aware of. The system almost certainly needs improving, but it is still the best we have available, and limiting its ‘alternatives’ remains the right thing to do as long as they fail to reach even its standards of objectivity.

Throughout my above arguments I have assumed the kind of science that conducts empirical experiments. Though those are certainly a relatively better way of testing theories than any other, that does not mean that it is not possible to similarly to come up with a theory through less concrete reasoning that more or less matches objective reality. I am quite willing to believe Berger’s and Luckmann’s theory as presented here is largely correct, which makes it an example. And as presented above, it is also an example of how a theory must go about justifying itself and negating rival views. This is quite right and proper. In fact, we so easily recognise this is what a theory needs to do in order to retain its believability, that even theories actually born through the unintentional development of the social view of the world end up doing this, as discussed especially in section 3. above. What we must do to obtain true knowledge is simple enough to state, though hard to achieve: we must come up with reasoning and proof that will prove some theory rather than only pretend to do so. We must detach ourselves from any initial point of view that has not originally been arrived at through explicit, rational means, and find one that actually fits the question at hand. This way we can theoretically reach objective knowledge, even if it is an endless task of refining one’s theories in practice. Such knowledge will have the same structure as part of a constructed reality as less true objectified knowledge, and will involve the same form of legitimation, but it will differ in being based on the non-human, non-social, outside reality as much as anything, and thus giving more or less reliable predictions of the world. Some curious minds have been trying this since ancient Greece at the very least, and in the time since then we have honed our methods so that now we may perhaps claim reliability in many cases.


This article is a slightly modified version of a paper with the same title written for the course “Philosophy of History” in the University of Turku in 2010.


Berger, Peter L. and Luckmann, Thomas 1991 [1966]. The Social Construction of Reality. Penguin Books. Referenced with bare page numbers throughout this essay.

(A note on footnotes: I imported them from an original text document and they work rather stupidly so that when you click on the number in the text, it will take you down to where the corresponding number is at the top of the screen, but since there is an extra bar at the top of the blog page, the number — which can be clicked to get back — will be hidden behind that. My apologies, but I’m not redoing them all totally. Most of them just give page numbers anyway.)
















“[E]mpirically, of course, these levels overlap” (112).








127, see also 128-30.






105, emphases in the original.






Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s