I want to assume now that you have built a network, internalized most of the lessons that I have outlined above, and generally gotten yourself established in your field, and that you now want to understand the dynamics of the research world in a deeper way.
Everybody that you work with has been building their own network in more or less the same way that you have been, and the institutions of research create tremendous incentives for everyone to keep on doing so. Beyond that, however, the institutions create new incentives for people who have reached the plateau that you have — providing yourself with a functioning network — and who want to move on to higher levels of accomplishment in their field. To understand these higher-level career strategies and their implications, another round of basic concepts will be required. This section lays out these concepts seriatim, and the next section agonizes over their moral consequences.
I should warn you that some of these concepts concern the more dysfunctional aspects of research institutions. My purpose in explaining these concepts is not to demoralize you, but quite the contrary to help you notice them, avoid them, if necessary defend or even cure yourself from them, and ultimately adopt a bemused distance from them as you go about the daily work of positive community-building.
The invisible college
When most people look at the university, they see a physical campus with buildings and people. Even if they went to college themselves, they probably have little understanding of the institutions of research. Accordingly, as you are socialized into those institutions yourself, you will probably acquire a different awareness of them. You will develop a professional network that includes researchers at several universities, and you will learn about the people and activities at those other universities. As a result, you will acquire a mental map of numerous universities and their associated orientations, reputations, and histories. This map will be very real to you, and you will probably know more about your counterparts in a university on another continent than you know about the people in the building next door to you. The interconnected global research network is largely invisible to outsiders, and for that reason it is called the “invisible college”, a term that derives from Diana Crane’s book “Invisible Colleges: Diffusion of Knowledge in Scientific Communities” (University of Chicago Press, 1972).
The concept of an invisible college is useful for several reasons. First of all, it helps to explain some of the institutional tensions that universities face: individual researchers generally identify more strongly with their invisible college than they do with the organization that employs them. After all, it is principally the invisible college that evaluates the researcher’s work by writing letters and refereeing articles. Universities are always threatening to be pulled apart by these centrifugal forces. Industrial labs, likewise, often have trouble persuading researchers to focus on the issues that affect customers, because the researcher’s long-term career success depends on staying current with research agendas in the invisible college.
Invisible colleges also help explain the emerging uses of technology in research. “Collaboratories”, for example, are on-line research community environments that cause invisible colleges to become, so to speak, more real. Most invisible colleges already have conferences, journals, and the like. They may even have Web sites and mailing lists. In each case the pressure is toward ever-greater integration of the different research groups within an invisible college. As the collaboratories become more technically feasible, these pressures will become even more intense. Ongoing real-time collaborations between researchers at different sites will become more common, and seminars might even be held at several sites simultaneously over video links. The details will depend on the needs and finances of each field, of course, but the general direction of the pressure toward integration will be largely the same. It is worth wondering, then, whether too much integration can be a bad thing. It is useful for each university to have its own distinctive approach to a field. Diversity is good, and the institution only supports diversity if a new approaches can colonize a small number of universities without excessive pressures to be interlocked with their opposite numbers at other universities. This may be an important issue in the future.
Finally, the concept of an invisible college helps keep you human. You can become so immersed in your own particular invisible college that you become oblivious to your environment — the neighborhood where you work. Think, for example, about the other universities in your region. Are they on your map at all? Do you feel bad about that?
Let us take the concept of an invisible college a step further. Imagine a vast diagram of all the professional networks in the world of research. In this diagram, everyone will be connected to everyone they know. Abstract as it sounds, such a diagram can actually be drawn with reasonable accuracy by following the citations in the researchers’ published work. The analysis of these citation links is called “bibliometrics”, and is a scholarly industry in itself. Throughout this article, I have been painting a picture of the structure of these relationships. When two researchers have become members of one another’s professional networks, they maintain a sort of surveillance of one another. They read one another’s work, monitor one another’s career progress, hear reports on one another through common acquaintances, update one another in periodic conversations at conferences, and so on. Their relationship has an architecture — a structure and logic that are dictated largely by the workings of research as an institution.
On one level, the architecture of relationships in the research world has not changed much since the Renaissance. Scholars have always read each other’s work, corresponded, traveled to visit one another, cooperated and competed, and so on. So what has been changing in the world of the Internet, not to mention cellular telephones, cheap air travel, and other technological advances? Those new technologies do not change anything on their own, but they do provide tools that people use to do more of the things that they already want to do. The institutions of research create tremendous incentives to keep in touch with the other members of your professional network, and that’s what’s happened: people are in much denser and more continuous contact with their professional contacts than ever before. It is only a slight exaggeration to say that we’re heading toward a world in which everyone is a constant presence for everyone else. Technologies that are currently under development will propel this trend even further. Digital libraries, for example, will allow everyone to monitor everyone else’s publications in real time, and cheap, high-quality video links will make it possible to organize seminars at a distance. While they will not eliminate face-to-face interaction altogether, these technologies will allow researchers to maintain even more continual contact than they do today.
This development is striking, and it counts as a new chapter in the history of the human person. Barry Wellman calls it “networked individualism”. Networked individuals (such as yourself) are like air traffic controllers who, by using a video display and audio communications, constantly maintain a mental map of all the planes in their airspace. This effect can be quite tangible when you are reading your daily e-mail, and it can be especially tangible when you are working on a large-scale professional project, like organizing a conference, that requires you to keep track of the status of dozens or hundreds of individuals, or to reach out selectively into the space of individuals in your field to identify the best speakers, authors, referees, or meeting participants for a given purpose. As the world becomes networked, you will have to decide consciously how to manage the blizzard of communications that your network will entail.
The expanding universe
So far I’ve been making it sound as though all networks are equally good. Start with what you care about, get some research going, and build a professional community for yourself around that research topic. And that is still my advice. Nonetheless, the problem of building a network takes on another dimension when you adopt a longer-term view. If you are entering the research community at the usual age, just out of college or a few years afterward, then you have a whole career ahead of you. To really prosper, and to really be part of something exciting, you want to join a field that is growing — what I call an “expanding universe”. A field that is shrinking is generally an unhappy place to be; at best it spends its time negotiating mergers and acquisitions with other shrinking fields, hoping to maintain the critical mass that is necessary to be a viable political force. People find themselves fighting over fewer resources, and they have a much harder time attracting new blood. A growing field, by contrast, easily attracts new people. And resources are multiplying, so people don’t need to fight one another. Instead they can join together in the collective enterprise of laying claim to the new territory that is opening up.
Smart students have a powerful instinct for expanding universes, and often spot them before the famous people do. How the smart students work this trick is one of the deeper mysteries of the professional world: after all, they are just students, and thus lack the extensive networks that are normally required to see big patterns. Part of the answer is simply that they are young. The way that ideas change is that the people who believed the old ideas die (Planck said this), and in that sense intellectual trends are driven by the interests of the young. This is one reason why it is okay for you to pursue the research that you personally find exciting: despite all of your unique individuality, you are also a product of a place and time, and even before you start networking you can be confident that plenty of other people will develop research interests that are more or less on the same wavelength as yours. You will network with those people, and when the old people die you and your cohort will inherit the world. At the same time, you can frame your topic in a lot of different ways, and it’s helpful to frame your topic in a way that other people can relate to. That is part of what I mean by “articulating an emerging theme”. In trying to articulate the theme that unifies the research of your peers, and that puts a name on what they find exciting about their research, you will be intuiting — indeed, creating — the expanding universe of your generation of researchers.
You are probably familiar with the general idea. Negative feedback is when forces operate to keep a system in equilibrium, pushing it back toward its nominal value whenever it drifts away. Positive feedback, by contrast, amplifies small disturbances so that they feed upon themselves and become ever greater. Complex real-world situations generally combine kinds of feedback, but it is useful to consider some of the positive feedbacks that promote successful careers. Let us say that you happen to mention topic X in a speech, and a reporter calls you to comment on it. You may not be an absolute authority on X, but if you are the first person to be quoted on X then you need to start studying. Why? Because reporters often decide who to call for quotes by looking in Nexis and seeing who has been quoted in earlier articles. Having been the first to be quoted, you will also be the second, third, and fourth. Soon you will be closely identified with the issue; nobody else will have a chance. The same thing can happen in many other contexts, including speaking engagements, consulting jobs, referrals, and (to a lesser degree) citations. (Of course, once your position has become entrenched in this way, it’s not positive feedback any more. Now it’s negative feedback, as institutional forces operate to reinforce the status quo in your favor.)
Another type of positive feedback is learning: if you learn about an activity (such as a certain experimental procedure), then you are more likely to get further chances to engage in the activity, thereby learning some more, so that you eventually become a leading expert. Yet another type is networking: if everyone knows that you have a big professional network, then they are more likely to want to meet you, thus expanding your network. People often stumble into careers because these types of positive feedback get started by accident, and good career strategies always encourage positive feedback. Pick an emerging issue and stake it out as your own: become publicly identified with it, learn the details of it from practical experience, and build professional networks around it. If you pick a good issue then the universe around it will expand, and your career will expand along with it. Picking an emerging issue is like placing a bet; your own intellectual intuition is the best guide to the best issue, but internalizing the views of others through networking is a good way to deepen your intuition.
Positive feedback also applies to departments, universities, and industrial labs. People want to work with the best people in their field, and so whichever organization first gets a critical mass of strong people can hire the best, thus locking in its position over the long term. This fact helps explain the hiring strategies of deans. As a general matter, it is in the dean’s interest to build specific areas of strength that correspond to expanding disciplinary universes. That is, the dean’s job is to build a group of researchers in a field whose prominence and resource base is likely to grow in the coming years. Choosing a research area that represents an expanding universe is obviously a good strategy, because an investment in that area will pay off as the field becomes more prominent. But it is a good strategy for another reason, which is that existing, established research areas have already become dominated by other organizations. Those organizations benefit from positive feedback, and so it is little use to compete with them. Instead, the dean seeks to get positive feedback working from scratch in a new area.
Arbitrage is a concept from finance. An arbitrageur monitors two or more markets, looking for gaps in prices. If apples are selling for $1 uptown and $2 downtown, the arbitrageur will swoop in, buy some apples uptown, and sell them downtown. The resulting profit will depend on communication and transportation technologies, and the magnitude of the price gaps that open up in practice will be limited by the number of competing arbitrageurs. Fully arbitraged markets have uniform prices. To remain profitable, therefore, an arbitrageur must innovate technologically or search for markets that are not yet well arbitraged.
Something similar happens in the research world. A researcher might notice that a concept that is well-developed in one field can be applied to problems of wide interest in another field. If the concept is still unknown in that other field, then an arbitrage opportunity exists. This is how management consultants work: they work with one company, learning that company’s organizational and technological skills, and then they sell their enhanced skills to other companies. It is also the way that many careers are made in research: either by shifting a steady stream of concepts from field A to field B, or by taking a particular concept from field A and looking for many different fields where it can be applied, or by moving from one field to another, picking up concepts in each and then looking for another field where they can be applied. These are entirely honorable ways to make a living, and they provide the intellectual cross-fertilization that keeps fields healthy.
The position of the arbitrageur can be understood in terms of social networks. In some cases an arbitrageur can learn new concepts, or contribute to new fields, simply by reading books. More commonly, however, the arbitrageur builds a network in each field, consisting of those researchers whose work is relevant to the arbitrageur’s own interests in that particular field. Of course, the very idea that the arbitrageur builds two different networks is somewhat artificial: the whole research world is one single network that is more highly connected in some regions than in others. “Fields” can be identified by their ideas and methods, but they also correspond to regions of high interconnectivity in the sprawling network of the whole research world. Arbitrageurs are effectively taking advantage of regions where the networks are relatively thin, importing and exporting useful goods (ideas, concepts, methods, tools) based on a strong understanding of supply on one side and demand on the other.
The opportunities for arbitrage are one reason why I have encouraged you to ignore disciplinary boundaries as you build your professional network. By looking for professional friends who are related to your research interests in several different ways, I suggested, you would create a network that looks like spokes in a wheel, of which you are the hub. If some of these people have nothing in common with one another then that’s a good thing. It means that you will be able to establish a “trading zone” through which good ideas can transfer between fields that are not otherwise connected. By spanning several research communities, you will have more intellectual resources and career options than if you simply tried to join an existing group.
As a scholar, you are certainly aware of your responsibility to cite relevant work by other people, especially when your own work builds on it. Your papers, like most people’s, probably contain sections that are largely devoted to citing past work, and you probably distribute citations through the rest of your paper as well. This is good; it is part of the process of knitting yourself and your work into the web of relationships in your community. But you can also look at these citations another way: as a narrative of the history of the field. These narratives may not be great literature, but they are narratives nonetheless. They have characters, events, and a chronological story line. They recount the creation myths of the field, its stages of development, its conflicts, its heroes and villains, and so on. The narratives in your paper will be shaped by your reading and relationships, but they will also be influenced by the narratives that you have heard or read from others. It is fairly unusual, for example, for a scholar to come along and tell the history of a field in completely different terms, recognizing different founders or different heroes, or giving a central place to different innovations and departures than the ones that normally form the backbone of the field’s narratives of itself.
Where do these narratives come from? At one level, everyone fashions their own narrative, connecting the dots among the various prominent works that relate to their own. Having laboriously rehearsed their personal version of the field’s narrative in their dissertations, they keep it up to date as their own work evolves, and as new work appears. At another level, however, the narratives are constructed collectively. People who do related work will probably have related narratives, and people who work in the same field will probably derive much of the outline of their narrative from whoever founded it.
When someone founds a field, they are usually very concerned to give the field a proper history. This might involve identifying precursors, marking out the differences between the new field and older fields, making clear which work the new field defines itself against, and so on. Later on, other people in the field will be sure to cite the people who have most influenced them. Peer pressure will grow to cite particular works that are thought especially important.
Over time, a more or less conventional narrative will take form. This conventional narrative is not a simple thing. It may settle disputes over who should get credit for a given innovation. It may embody a collective judgement that certain works represented side branches or cul-de-sacs, and that certain other works represented the main line of development. Ideas from certain works will become part of the routinized story that people tell about their field, and those works will be heavily cited accordingly. Or a work may introduce an idea that seems revolutionary at first but then starts to seem so obvious that people forget that it needs to be cited any more. Some authors may make a special point of insisting that their work be cited, where other authors may not care as much, or may not be around to check up. In short, the conventional narrative emerges as a sort of collective negotiation among the field’s members. And as new scholars encounter the conventional narrative in their readings and lectures, it settles into place and becomes practically irreversible.
I mention these disciplinary narratives for several reasons. First of all, I don’t want you to be imprisoned by them. Look at them as narratives, as stories that are told according to certain conventions, and that could have been told differently. See their political character not necessarily as a sign of bad faith, simply as a sign of their having been created by human beings through their dealings with one another. As you read the literature, consider whether the conventional narrative of your discipline should be rewritten.
Declare independence by quietly citing works that have been unjustly neglected by others (such as works by people who haven’t done their networking). Ask yourself if the field’s founder constructed a creation myth that exaggerates its differences from what came before, or that emphasized a single moment of invention when in fact (as often happens) the basic ideas emerged in several places more or less at once. Maybe you want to rewrite the narrative a little bit in your next paper. And think about how your own work deserves to fit into the narrative. Describe your work accordingly in your papers, and do make sure that the people who should be citing you feel a bit of peer pressure. You don’t have to be a jerk about it, but you don’t have to get trampled either.
There is an aspect of disciplinary narratives that I want to emphasize in particular. When your field was originally founded, the founders probably overcame opposition from an existing establishment. As a result, the rhetoric that they developed and taught to their students was probably preoccupied with that particular fight. For example, artificial intelligence (in which I was trained, and whose story I will tell in more detail in a moment) began as a counterrevolution against behaviorism in psychology. Because of this the rhetoric of AI is saturated with turns of phrase that are designed to do two things:
(1) set up a cleanly defined opposition between AI and behaviorism, and (2) portray AI as right and behaviorism as wrong. The AI people won their fight with behaviorism, which hardly exists any more. And yet the fight goes on. The rhetoric of the field is still aimed at defeating behaviorism, and this causes AI people to interpret nearly any criticism as a resurgence of behaviorism, even when it clearly is not. It also causes dissidents within the field to reinvent behaviorism under one guise or another, simply because that is what’s thinkable within the vocabulary of the field.
This is the sad irony: even though AI won its fight with behaviorism, it did so by making itself much more similar to behaviorism than it should have. The problem is not so much with goal (2), portraying AI as right and behaviorism as wrong, as with goal (1), setting up a cleanly defined opposition between AI and behaviorism. In order to set up this clean opposition, it was necessary for the AI founders to commit themselves to many of behaviorism’s foundational assumptions, such as the idea that cognition takes the form of an input (stimulus) which causes something-or-other (a blank zone for the behaviorists, a cognitive process for the AI people), which then causes an output (response). This framework has not served AI especially well, for example because it distracts attention from the ways in which people and robots engage in complex activities that are embedded in complex environments. Yet this complaint is hard to express in the language of AI, whose organizing question is still, “is this behaviorism, and if not then what’s the problem?”. You can accomplish a great deal by spotting this sort of out-of-date controversy and deciding not to participate in it. Even in cases where the “enemy” establishment is still very much in force, you will accomplish much more by honestly digging into the strengths and weaknesses of the two polarized sides, looking for a synthesis rather than a fight, than you will by joining someone else’s ancient struggle.
Advisors’ incentives to stifle creativity
The next concept that you need is not so fun. This is the incentive that thesis advisors have to stifle the creativity of their students. It’s an insidious phenomenon, and it is not entirely the advisors’ fault. Here is how it works. Your advisor will organize seminars, or otherwise recommend reading, and the reading lists that result will derive from the advisor’s own voice — from an intellectual map of the world that reflects the advisor’s own effort to define a research program and situate it within an existing network of professional relationships. If you confine your reading to your thesis advisor’s recommendations — or, even worse, if you feel so overwhelmed with work that you accept your advisor’s interpretations of those readings rather than engaging with them afresh yourself — then your thinking will be organized and bounded by your advisor’s thinking. You will talk the way your advisor talks, cite the same work, address the same audience, and so on. Of course, this needn’t be a disaster. If you are smart, and if your advisor has chosen an expanding disciplinary universe, then you will write a good dissertation within that universe. You will get a good job, and you will take your place in a hierarchy. When the people in your advisor’s cohort finally retire, then you will be in charge. It is not such a bad life. But it is not the life that you were meant to live — the life that you would create for yourself if you complemented your advisor’s teaching with some autonomous learning of your own, driven by your own sense of intellectual excitement and your own intuition for the expanding universe that is taking form on completely different ground from your advisor’s.
So is your advisor deliberately brainwashing you in order to build an empire of clones and acolytes? Perhaps. Some advisors do this consciously, I am sad to say. It’s their way of proving to themselves (and, they think, to others) that they are a success. After all, they are evaluated on their “impact” in their fields, and one way to create the illusion of impact is to program your students so that they are forever citing your work. Perhaps they just want to make sure that they do not die forgotten. Or perhaps they simply get locked into a fixed idea about your thesis topic and try to “help” you graduate on time by keeping you narrowly focused on that topic. Of course they rationalize it in various ways. But with other advisors it happens inadvertently. Your advisor is not God, cannot read everything, and inevitably sees the world in particular ways. Your advisor lives in a world that seems very big, and if your field is expanding then you could perfectly well construct a world within that world that itself seems very big. The alternative is not to renounce your advisor, but simply to reach out and take a broader view.
I spoke of a growing field as an “expanding universe”, but what exactly does it mean to say that a field is growing? Of course, on one level a field grows when more people join it. But that doesn’t explain much. Nor does it explain much to say that a field grows when more money becomes available to fund its research, though money is surely not a trivial matter. At a more fundamental level, the size of a field is determined by the turf that it has staked out. My choice of the word “turf” is a little misleading, in that actual literal turf — geographic territory with grass on it — exists in a fixed quantity, so that the phrase “turf war” connotes a bloody, petty, zero-sum game. But that’s not what I mean here. In the research community, turf arises when an intellectual leader defines a research agenda
— that is, provides a rhetoric for articulating research topics, arguing their importance, and defending their legitimacy. Having been made researchable, those topics can now be turned into refereed journal papers, and thus into grant proposals, promotions, and careers.
Here are some examples. When Herbert Simon and his cohorts founded artificial intelligence in the 1950s, they created turf. In fact they created a huge amount of turf, since the general formula of using computational structures in analyzing human mental life can be applied in thousands of ways. Just pick a phenomenon of human mental life (remembering, planning, improvising, etc), select or devise a computational structure that seems generally analogous to it, build a computer program, and talk about the program in ways that make it seem similar to what people do. Other examples of research programs that create turf include Richard Posner’s revival in the 1970s of the economic analysis of law (pick a legal issue and apply the language of supply and demand to it) and Noam Chomsky’s founding in the 1950s of the modern study of syntactic analysis by means of formal language theory.
What’s really striking about the case of Chomsky is that his actual territory of research concerned some extremely narrow questions about the formal relationships between certain kinds of grammatical structures, for example when assertions (“John took a six- pack to the party”) become questions (“What did John take to the party?”). Even though these questions are tiny footnotes in the big picture of linguistics, Chomsky nonetheless managed to found an enormous research enterprise, one which many linguists have been brought up to regard as nearly the whole of the field. Chomsky was successful in founding such a large research program for a simple reason: formal language theory provides the intellectual tools to manufacture researchable topics. Accordingly, every paper in Chomskyan linguistics — including several subfields of linguistics that broke off from Chomsky’s own projects while retaining nearly all of the intellectual foundations that Chomsky created — is written according to a sort of grammar that Chomsky defined and institutionalized.
These examples point to the actual nature of turf. In order to do research, and in order to publish your research, you need a research topic. Turf is, in part, a method of manufacturing research topics, a formula for producing the raw material from which people make their careers. But turf must be defended. On a small scale, you can only publish your research if you can defend it to the satisfaction of the journal’s referees. And on a large scale, a research program depends for its funding and other resources on its reputation in the larger research community. It is important to distinguish here between two kinds of legitimacy that research needs. In a narrow sense, the claimed results must be seen to follow from the premises. But in a broad sense, the research topic itself must be seen as legitimate: that is, as novel, conceptually coherent, defensible in its working assumptions, intrinsically important, likely to lead to practical applications, likely to lead to more productive research, and so on. The precise criteria will depend to some degree on the field (engineering is evaluated differently from history), but every field needs someone to put up a fight when the legitimacy of the field’s research topics comes into question. And many topics require a great deal of defending, given that the many idealized assumptions, unmotivated choices, and unredeemed IOU’s they entail. That is what a visionary founder like Simon, Posner, or Chomsky does. These guys don’t just publish technical papers within their field — what Aristotle would call “esoteric” work, that is, work that is directed to the community of like-minded researchers within the field. They also publish “exoteric” work, that is, work written for a broad audience that explicates the field and defends it against critics, either explicitly, by answering the critics’ charges one after another, or implicitly, by providing the field with conceptual and rhetorical foundations that are meant to be understood by insiders and outsiders alike.
These sorts of exoteric apologetics for a field’s turf are one more important way in which people become dependent on their thesis advisors (or, indeed, on their advisors’ advisors). If you grow up intellectually within the small world of a particular field, you will never be called on to defend the legitimacy of your research topic. You will probably read the founder’s exoteric texts, and you will learn to talk the field’s rhetoric, but you will probably not have occasion to really internalize the arguments of the field’s opponents.
Many people reach mid-career in this position, and I believe that it induces in them a kind of vertigo: they have staked their careers on the continued viability of a chunk of turf that they did not create and cannot defend, and if that turf loses its legitimacy then their careers will evaporate. It takes a lot of reading and networking to establish yourself in a different research community than the one you were trained in, and it’s especially hard if your training has not encouraged you to develop a robust intellectual life outside the fine details of your particular lab’s research program.
This is, in my opinion, a major cause of some of the less fortunate cultural phenomena of research world, including us-and-them stereotyping of other research communities and a tendency to make a virtue of narrowness or to overinflate the real scope and potential of the field as it stands. I believe it also explains the fury with which many researchers respond to any criticism of the foundations of their research enterprise. On one level, the organizing ideology of the turf routinely caricatures opponents as irrational, unscientific, etc, so that critics are heard to be saying things that are literally crazy. After all, most people’s fields seem like whole universes to them, and their networking is often confined to people who share the same ideology. On another level, the researchers themselves are unconsciously terrified that their careers will explode if the criticism succeeds. As a result, they are motivated to exaggerate the extent to which the real concrete results of the research program have established the truth of the intellectual school within which they work. These pathologies are not universal, of course, and they vary greatly in their intensity. By describing them, I want to help you identify them and avoid falling into them. People in the research world are too often honored in proportion to the amount of turf they created, and not in proportion to their intelligence and goodness. If you can shake off this bad habit then you can start honoring the right people, and honoring them in the right way.
When you are a student, you tend to take for granted the whole institutional framework that you are being socialized into. You might complain about it, and you might even spin conspiracy theories about it, but you do not have the information that would be necessary to understand what the institution really is and how it really works. This article is intended as part of the solution to that problem. Having explained how research institutions work on a micro level, then, I want to explain where they come from. Let’s take a simple case: a workshop. You could start a workshop yourself; I explained the procedure back in Section 3 under the heading of “intellectual leadership”. If you can round up a critical mass of attendees, then you just do it. Your workshop meets, everyone is happy, and the idea circulates of maybe doing it again next year. Maybe someone else takes the lead, hosts it in their own seminar room, and so on. If enough people keep on feeling like the workshop is worth their time, then maybe it takes on a life of its own. If your emerging theme happens to define an expanding universe, then your workshop will grow. Twenty people might attend the first year’s meeting, forty people the next year, a hundred the year after that, five hundred the year after that, and so on. At that point you probably call it a “conference”, and maybe you and the other central ringleaders organize yourself into some sort of standing committee. Maybe you start an organization, a mailing list, or whatever you build consensus around. Your emerging theme has been institutionalized.
Notice something important: at no point did you have to ask anybody for permission. You just did it. It’s a free country, so you used your freedom of association to associate with other researchers who share your interests. You can accomplish a great deal this way. But this is just the beginning of the story. The next step comes when you start a refereed journal. This is something else that you just do, although now you need to persuade a publisher that a critical mass of interest exists to make the journal work as a business. You start a journal in exactly the same way that you organize a workshop: having already built a network, organized some workshops or conferences, etc, you circulate a draft proposal to the ten people whose names everyone would most expect to see on the journal’s board of editors, and if you feel like the proposal has some energy behind it then you go ahead. The people in your network will welcome the new journal because they don’t feel their papers are being refereed fairly by the existing journals; your journal will ensure that the referees, while presumably maintaining high standards, will at least comprehend the papers and thus be able to judge them fairly.
Above all, a journal gathers up and organizes a community of people who share a complex of research problems, so that everyone can at least anticipate that the referees will regard their problems as legitimate, even if they do not agree with the details of the research itself. That is the service a journal provides to its contributors. In this way, a journal frees its contributors to write the papers that they really want to write, and it ensures that their vitae will now fill up with bona fide refereed journal articles. Of course, the value of those vita entries will depend on the larger community’s evaluation of the quality of research in the journal, and for that purpose it will still be necessary for the emerging field’s founders to engage in the exoteric apologetics that I explained above.
Even so, a journal does a tremendous service to a community by enabling its members to be the judges of one another’s work.
A journal is a step on the road to institutionalization, but it still does not explain where resources come from. Let us consider one type of resource, probably the most important of all: job positions at universities and other research organizations. (These job positions are often regrettably called FTE’s, for “full-time equivalents”, since it is somewhat common for someone to be appointed half-time in each of two different departments.) Where do job positions come from? In the case of universities, they come proximally from deans. When resources become available at the overall university level, the deans engage in politely savage warfare to lay claim to some new positions, and then they allocate the new positions among the departments in their domain. Meanwhile, each department tries to help the dean by describing in compelling terms the turf that is opening up at the cutting edge of their field. As your research area becomes institutionalized, your collective job is to define the emerging turf you see ahead of you, to make the case for this turf seem compelling to your dean, and to help your dean make the turf seem compelling to the university hierarchy. That is where job positions come from. The process is not always sweetness and light, because much of it takes place in committees, different subsets of whose participants are angling to define their own research area as the Next Big Thing for the allocation of job slots. An ascendant field can find itself laying claim to literally dozens of job slots in a period of a few years, and at each step the field’s members will be working their networks furiously to produce a steady stream of high-quality candidates for the jobs. Maybe this is how you got your own job.
Routinization of charisma
Now that I’ve explained how to institutionalize the new research area you’ve founded, notice something important about institutions: they arise through individual initiative. This fact has many consequences. When you begin your career in research, you will encounter a landscape of already-established institutions — they will be called “fields”, “journals”, “conferences”, “agencies”, and so on — and every one of those institutions will have arisen through exactly the same kinds of individual initiative that I am recommending to you. Someone built a network, articulated an emerging theme, organized people around it, connected the emerging constituency with a supply of resources, and created new organizations. Those new organizations then settled down and took on a life of their own.
Institutionalization thus entails a process of maturation: from the initiative of a founder to the more anonymous settled patterns of the long term. Max Weber called this process “the routinization of charisma”. The founder acts as a kind of enterpreneur, articulating a discourse for the field and creating turf within which others can pursue their careers. As a result, institutions often retain the fingerprints of their founders. If the founder’s overwhelming imperative was to defeat an existing establishment, that imperative will probably continue to structure the field below the surface. If the founder’s overwhelming imperative was to secure the patronage of military funding agencies, then research problems will probably continue to be framed in that way after the founder is gone — even if nobody realizes it. If the founder was a poor organizer or had a personal preference for a chaotic institutional style, then that style may persist for decades afterward. Nothing is inevitable, of course, but institutional patterns do tend to persist once they are put in place. And these patterns originate with the founder, and with the opportunities and challenges that the founder originally confronted.
What does the routinization of charisma mean for your own career? Several things. (1) Don’t be fooled by the sense of permanence that every institution projects. Your field’s founder — and especially the founder’s students — codified a fragmentary mess of ideas into survey papers, syllabi, and textbooks, all of which are supposed to look seamless. They’re not seamless, though, and you should assume that everything is much less stable and coherent than it appears. (2) To find the seams, you should study the history of your field. Go back to the founding documents, and get old-timers talking on social occasions. Understand the context in which the field was founded, and look for left-over patterns that are no longer relevant in the present day. It’s alright to have respect for founders, but realize that they are mere human beings, products of the times and places in which they lived. (3) When you do discover these obsolete patterns, deprogram yourself. You will inevitably have ingested a sprawling network of unarticulated assumptions into your own thought patterns, and if you can liberate your mind then you can improve your research.
(4) When you start creating institutions yourself, be responsible. You don’t want your personal quirks — or your short-term opportunism, rivalries, and greed — to be transformed into settled canons that get taught to generations of unsuspecting students.
In order to grow, a research community must create more turf — that is, broader and broader territories of legitimately researchable topics that the community’s members can publish on. Because turf is not a fixed quantity, it is often possible to create new turf within the existing boundaries of the field. In this sense, turf is “nested”, meaning that people build their careers and reputations by mapping out territories of researchable research topics within the broader continent that the field’s founder(s) had already mapped. Thus, for example, the turf of artificial intelligence, having been mapped out in a general way by Herbert Simon and others of his cohort, subsequently developed well- defined subterritories, such as AI subfields of “planning” and “machine learning”. In each case, a student or student-of-a-student of one of the field’s founders made their career by institutionalizing the new subfield: articulating an emerging theme, building a network around it, organizing meetings and journals for the network’s members, and so on. There is nothing wrong with this, of course, if it’s done well. I just want you to see the pattern. Similar things have happened in many other fields.
Of course, I do not mean to suggest that the process is mechanical, or that its success is guaranteed. The AI subfields of “vision” and “robotics” for example, ended up being institutionally outside the AI community, simply because the people who were doing well-regarded work in those areas were largely situated within other communities, such as neuroscience and mechanical engineering. So the boundaries between different research communities are variable.
And that brings me to the concept at hand: imperialism. One way that fields create new turf is by applying their organizing concepts and methods to subject matters that have historically been the “property” of other fields. Economists of the dominant neoclassical school, for example, have gone to great lengths to portray all phenomena of human life, from education to child-rearing to the fine details of individual cognition, as examples of neoclassical ideas about economic rationality and allocative efficiency. Scholars in the fields that they have invaded, sociologists for example, are often horrified by the strange and extreme violence that the economists’ formalizations appear to inflict on their proprietary subject matters, and much gnashing of teeth has ensued, together with some genuine attempts to build bridges. Given the workings of the institutions, however, sniping at the invaders does little good. So long as they can institutionalize themselves, establishing organized research communities whose members are called upon to evaluate one another’s work, external criticism must be taken to the larger and slower court of public debate and institutional review.
Economics is probably the most imperialistic of all research fields, but the process is nearly universal. Careers require turf, and they require coherent communities that can collectively defend their turf. Research communities therefore try continually to apply their overall “story” to new subject matters. These campaigns can lead to faction and warfare. Movements can develop pathological ideologies to justify their imperialism, in extreme cases labeling other work as “old-fashioned” or denigrating any allocation of resources to others as “lowering standards”. Other cases are not pathological at all, and simply represent healthy competition. The dynamics are complicated, and they are hard to see, except from the standpoint of the highly-networked individuals who staff academic hierarchies and sophisticated funding agencies. To watch them happening, however, all you really need are some basic concepts (like the ones I am explaining here) and the disposition to build far-flung professional networks.
As the last several paragraphs suggest, political life in the research world consists largely of contention for resources among various research communities. To get some perspective on these political processes, it helps to understand the phenomenon that anthropologists call segmentary politics. Let us imagine for the moment (simplifying greatly) that society is organized hierarchically: households belong to neighborhoods, which belong to towns, which belong to regions, which belong to nations, which belong to broad cultural groups. If you look carefully at a town, you might find that the people in adjacent neighborhoods are continually struggling with one another. But if one town attacks another, those conflicts might be cast aside as everyone rallies to the defense of their town. Rivalries among towns might likewise subside as tensions arise between regions, but these tensions might dissolve temporarily when war threatens between nations, and so on. Each element of this picture — a household, neighborhood, town, region, nation, or broad cultural group — is called a “segment”, and segmentary politics consists of an endless negotiation of conflict and solidarity among adjacent segments. Of course, the picture can be more complicated when groupings cut across borders, for example in former colonial areas where national borders are randomly related to the
borders among cultural groups. But even the simple picture is useful as an antidote to the even simpler picture of undivided national loyalties.
As you start participating in the institutional life of your field, you will probably notice segmentary politics yourself. Individual members of a department may dislike one another, but they may find it in their interest to remain allies in the department’s factional struggles. Those factional struggles may in turn be put aside when the department as a whole is threatened in some way, or when various departments are making their case for increased resources to the dean. Within the invisible college of a research community, likewise, segmentary politics might motivate opposed tendencies to pull together long enough to present a unified front to a funding agency that is thinking of starting a new program in their area. And the different schools of thought within a discipline might organize to fend off imperalistic assaults from other disciplines around them.
Segmentary politics is distasteful. In describing it, I certainly do not mean to praise it. You will probably find yourself engaging in it for self-defense if nothing else. But your real job, in the long term, is to transcend it. That is what networking is for. The broader and stronger your network, the less subject you are to the randomness of people’s local rivalries. This is one reason why I have editorialized here against disciplinary bigotry: the too-common stereotyping of one discipline by another. The argument against bigotry can be generalized: the conflicts at every level of the segmentary hierarchy are usually organized by stereotypes that have arisen over years, if not decades. Some of these stereotypes may perhaps be justified: despite all of your own best efforts to identify shared values with people in various fields, there may remain certain intellectual orientations that seem completely worthless to you. But at least these will be considered opinions, or as considered as you can make them, and not the uncritical acceptance of other people’s stereotypes. And you should remain open, looking for previously unsuspected points of intellectual or moral contact with fields that have previously seemed alien to me. By maintaining this attitude of openness, you can avoid mindlessly closing yourself off from potentially new and constructive directions of networking.
The role of rational debate
I’ve been talking about “politics”, but many people at the beginning of their careers wonder what “politics” means. After all, many people equate politics with corruption, and they feel as though engaging in politics means instant damnation. It is important, therefore, to get a positive conception of politics. A good place to begin is with the role of rational debate. To make the problem concrete, let us imagine a faculty meeting where decisions are being made about which candidate to hire for a job. Most such meetings are conducted according to a formal rulebook such as Robert’s Rules of Order that gives everyone a chance to make motions, offer arguments, call for votes, and so on. In that sense everyone is formally equal. But we all know that other things are going on behind the scenes. People come into the meeting with their agendas, their coalitions, and so on.
What, one may well ask, is the purpose of holding a rational-looking debate, when the fix is probably already in?
Implicit in this way of asking the question are two stories. On the first, naive story, everyone has an open mind and wants only the best for the school and its students. The naive story suggests that the meeting will be a shared, rational inquiry into the outcome that is best for everyone. On the second, cynical story, everyone is angling for power. The cynical story suggests that rational-looking debate is purely for show, and that the outcome is already set. In fact, the reality varies a great deal, and is almost always a mixture of the two stories. A healthy academic department will be somewhat fluid in its politics, neither strictly hierarchical nor rigidly factionalized, but consisting of a shifting map of different groupings who see things in different ways, and who need to remain on good terms with one another because future issues may require them to form unexpected alliances. In that environment, rational debate does serve a purpose: most issues will have their undecided swing votes, and whoever communicates best with those swing voters will win. Of course, not every department is perfectly healthy, and human beings will always have failings. If you are on the losing side of a harsh political division, then your goal should be to leave, and networking is the best way to go about that.
Whatever the case, you should understand that “politics”, whether naive or cynical, always starts with the commonalities that people have established in their discussions with one another. If you don’t want to become enmeshed in “politics”, in the negative sense of that word, then you should cultivate the skill of identifying points of intellectual overlap with other people. Don’t let your pride get in the way by defining your intellectual agenda in one inflexible way. I’m not asking you to compromise your values, or to pretend to believe things that you really don’t. There are lots of ways of explaining various aspects of your intellectual interests, and you will navigate in political space much more readily if you decide to articulate the versions of your interests that establish points of contact with particular people. Once you do this, the dichotomy between the naive and cynical pictures of politics starts to break down: you will make common cause with people in an honest way, and they will understand that you are on their side.
Political coalitions will emerge in a natural way, and you will be in the middle of them. I will discuss these articulation of commonalities further in the next section.
I want to start the transition to the final section on ethical issues by making explicit a concept that has been implicit at many points along the way. In your new career as a researcher, you are entering a complicated set of institutions whose participants occupy a variety of different positions. You have to deal with these people, and you will get things done by building consensus with them around ideas and projects that you find important. Because these people have different positions than yours — department chair, dean, funding agency program manager, PhD student, journal editor, etc — communication with them will not be automatic.
Here is a way to think about it. Everyone has stuff going on in their heads — questions, concerns, agendas, precedents, peer pressure, and so on — and you can communicate with people better if you understand how they work. Partly this means understanding them as individuals, with their own backgrounds, histories, ideas, peculiarities, whatever. So if
you’re going to talk to someone, and something important is at stake, you should try to talk to someone else who knows them first. That said, though, much of what’s going on inside people depends not on their personalities but on their positions. Every dean has certain concerns, and so does every PhD student, every department chair, and so on. If you are talking to someone whose shoes you’ve walked in, a PhD student for example, then you have a small chance of empathizing with their concerns (assuming that you haven’t forgotten entirely what it was like). But if you are talking to someone whose world you cannot imagine, such as a dean, then you are likely to make mistakes.
The concept you need is reverse engineering: figuring out what the people are likely to care about, and then speaking to that. I am not talking about manipulation, and I am not talking about telling people what they want to hear. The point, rather, is to anticipate what issues they people will have, and to make sure that what you’re proposing takes those issues into account. Here is an example. In Section 6 on job-hunting above, I described one way to write a letter of recommendation: gather all the good things you can think to say about someone, and organize those good things into a coherent story. That’s a pretty good formula. But a better formula is to start from the concerns of the people who are likely to be reading the letter. universities and other research institutions are full of evaluation processes, and central to all of them is a dilemma: people are made to prepare defensible evaluations of research in fields where they have little or no expertise.
Everybody realizes that this is a problem, and almost everybody is responsible about reaching beyond their own knowledge. That is what letters of recommendation are for: they are evaluations from people who know the work. Yet those letters don’t fully solve the problem. Someone has to interpret the letters and convert them into up-or-down decisions that they can justify. What to do?
The major idiom for these justifications, it turns out, is comparison. Is this person the best in their field? How do they compare to other people at similar levels of advancement in the field? How does this program stand in some magazine’s reputational rankings? That is how people like CAP (the Committee on Academic Personnel that I mentioned above) will be thinking. A letter-writer, therefore, is well-advised to speak that same language.
An expert recommendation letter will say things like, “In preparing this letter, I conducted an informal exercise in which I assembled a list of several other prominent scholars at this person’s level of advancement, and in doing so I found that this person’s work is clearly ahead of the pack in terms of this, that, or the other thing”. That’s reverse engineering.
The concept obviously generalizes. In various sections above, I have explained what various sorts of people are worried about. Department chairs, for example, have to get courses and committees staffed. PhD students on the sidelines of faculty hiring want to get faculty hired who can teach the courses they want to take. In each case, you can accomplish a great deal by showing that you are aware of these oncerns. You don’t have to make them your responsibility, but you should not try to evade them or get around them. If someone has a valid concern, then you should get that concern on the table and cooperate in addressing it. The situation is harder, of course, when you’re dealing with someone whose concerns are not valid, or who is presenting concerns as valid that are
actually disguises for other agendas. In those cases all you can do is put your own valid concerns on the table and negotiate. But don’t get into the habit of acting like everyone else’s concerns are just negotiating positions. That’s not a healthy way to live.
Demographic tidal waves
When we speak of the university as an “institution”, the suggestion is that things stay pretty much the same. But the institution does change, and you need to understand how. One source of change is information technology, but a much more important source of change is demographics. Let me just consider the United States. After World War II, a huge number of soldiers came home from the war and went to college on the GI Bill. The university system expanded tremendously to accommodate this new wave of students.
Those same students also had a tremendous number of children, the famous “Baby Boom”, and when those children grew up, the university system kept expanding to accommodate them as well. The federal government’s research establishment grew explosively at the same time, and much of this money went to universities. These generations — the GI Bill and the Baby Boom — took the concept of an expanding universe for granted. The job market was so strong that they could take chances with their careers. Beginning in the 1980s, however, that picture changed. Student numbers stopped growing dramatically, and the academic job market contracted accordingly. Qualified college teachers were so numerous that many colleges moved away from tenure-track faculty and toward exploited part-timers. Academia started getting a bad reputation as a career choice.
That situation is about to change, for two reasons: (1) the Baby Boom generation is going to retire, and (2) the Baby Boom’s grandchildren are headed to college. The numbers in each case are overwhelming. The University of California, for example, plans to hire about 7,500 new faculty members in the next ten years — which is more faculty than it employs right now. In fact, higher education planners assert that it will be physically impossible to accommodate all the students who will be ready for college over the next decade. Despite what you have heard, therefore, this is the best time in human history to be entering the research field. The networking skills that I have been presenting are crucial when the job market is scarce, because the Baby Boom generation’s easygoing career strategies certainly don’t suffice. But networking skills will also be useful during the good times ahead, when the university will be completely remade in a short period.