Berger and Luckmann want to show that reality is socially constructed. This can mean two things: that the objective features of the world have assumed the shape they have as a result of social action; and the features of the objective world can only be understood through one or another conceptual schemes that are both incommensurable and irrefutable. What they actually show pertains to the first interpretation, not the second. The most enduring contribution they make is to work out the case for this proposition: We as persons, and the social relations and processes within which we act, are iteratively created by previous social processes and individual actions. So the book isn't about knowledge; it's about social reality.
It is apparent from page 1, that Berger and Luckmann have a non-standard conception of "knowledge". They define knowledge as "the certainty that phenomena are real and that they possess specific characteristics." This definition has more to do with the degree of subjective confidence that persons have in their beliefs, and less to do with the nature of those beliefs themselves. And yet the topic of knowledge, whether philosophical or sociological, is really only interesting if it sheds light on the ways in which cognitive entities arrive at and formulate representations of the world around them.
Berger and Luckmann want the sociology of knowledge to focus on commonsensical beliefs, not specialist or scientific knowledge. "The sociology of knowledge must first of all concern itself with what people 'know' as 'reality' in their everyday, non- or pre-theoretical lives" (15). This is a perfectly legitimate point. But it doesn't erase the need for conceptual analysis: what is the structure of commonsense knowing? How do ordinary people "parse" their daily experiences into an organized representation of their worlds? These questions are just a much of a philosophical issue for commonsensical knowledge as they are for Kant in his consideration of all empirical knowledge.
Much of the social world that we confront and about which we form beliefs has to do with institutions. Berger and Luckmann have a particular and narrow definition of an institution. "Institutionalization occurs whenever there is a reciprocal typification of habitualized actions by types of actors. Put differently, any such typification is an institution" (53). The examples they give of institutions are practices that have grown up organically -- e.g. conventionalized ways that a traditional village may have come to have organized the annual stag hunt. But it would seem that there are many things that we would call "institutions" that fall outside this paradigm. For example, the Internal Revenue Service is an institution. It consists of hundreds of thousands of employees, organized by a set of rules, disciplinary processes, and oversight mechanisms. It is true that this institution specifies regular forms of conduct by the various people who are part of the institution. But it certainly didn't come about as the "sedimentation" of simpler forms of practice. More generally, their definition of an institution doesn't seem to do justice to the social realities represented by organizations.
That said, this definition of an institution gives concrete meaning to one sense in which Berger and Luckmann mean to say that "social reality is a construction": the institution itself is socially created by a group of people. "It is important to keep in mind that the objectivity of the institutional world, however massive it may appear to the individual, is a humanly produced, constructed objectivity" (60). This is certainly correct. But it doesn't support or convey the other important implication of "social construction" -- the idea that the world we experience is fundamentally constructed in terms of the concepts that we impose upon it. This is the sense implied by Whorf and other conceptual relativists; but it doesn't find expression in B-L's analysis.
My overall assessment of the arguments offered by Berger and Luckmann here is somewhat negative: I don't think they are offering a "sociology of knowledge" at all. Instead, they are offering an interpretation of the actor (constituted by processes of socialization before biology is even completely finished) and of the social world in which we act (created by the practices, actions, and habits of concrete human beings over time). It is essentially a sociological theory of the actor-in-social-context. The discussions of primary and secondary socialization are empirically useful, in that they help steer us towards the concrete situations through which individuals learn about the roles and values they "should" recognize. Seen from that perspective, the key chapters (II and III) are interesting and helpful.
But the book has very little to do with the problem of mental representation; and it doesn't have much to say about social cognition. And the recurring theme, that there are alternative social realities, needs to be understood as relating to the social-stuff side rather than the knowledge side: social relations, habits, and patterns of social behavior could have unfolded differently. There is nothing inevitable about the specific forms of interaction "our" society has codified. We could have created different institutions. But given the institutions and practices we've got, the task of knowledge is determined: we need to discover through participation and practice how they work.
So I'm not too excited about this book as one that contributes to a better understanding of cognition -- I don't find Berger and Luckmann's analysis of knowledge and the social world very helpful. The problem is, that they don't have anything like a nuanced analysis of the relationship between thought and the world: the nature of conceptual schemes, the relationship between concepts and observations, and something like a naturalized analysis of evidence and belief acceptance. In other words, they aren't doing enough of the philosophical work that is needed in order to have a genuinely insightful basis for talking about the social construction of beliefs. We need to know what goes into beliefs about the world before we can get very specific about how those belief systems are socially conditioned or constructed. They acknowledge this limitation of their approach:
We therefore exclude from the sociology of knowledge the epistemological and methodological problems that bothered both of its major originators. By virtue of this exclusion we are setting ourselves apart from both Scheler's and Mannheim's conception of the discipline, and from the later sociologists of knowledge (notably those with a neo-positivist orientation) who shared the conception in this respect. (14)They don't seem to think this avoidance of philosophical issues reduces their ability to shed light on the topic. Unfortunately, I think they were mistaken.
In fairness, I should acknowledge that the kind of analysis I'm looking for isn't wholly absent. Here is a statement that comes closer to the kind of analysis that I find generally lacking in their book:
I apprehend the reality of everyday life as an ordered reality. Its phenomena are prearranged in patterns that seem to be independent of my apprehension of them and that impose themselves upon the latter. The reality of everyday life appears already objectified, that is, constituted by an order of objects that have been designated as objects before my appearance on the scene. … In this manner language marks the co-ordinates of my life in society and fills that life with meaningful objects. (21)This is the beginnings of a philosophy of knowledge. It provides place-holders for some of the chief aspects of cognitive representation: the identification of permanent "objects", a field of inter-related objects and relations, and a language in terms of which these items are represented and in terms of which one's beliefs about them can be formulated. Or in other words: this paragraph postulates concepts, a conceptual system, and an intensional orientation of the subject towards the world (applying language to the "objects" around him or herself). And this world is "intersubjective" -- other people share concepts and language with me, and are in similar relationships of interaction with the stuff of the world we inhabit (22).
So we have a start on the more conceptual side of the problem. Unfortunately, this strand of thought is not further developed throughout the rest of the book.