'A student assigned ... these textbooks might think that "real" archaeology is an exotic excavation in the jungles of Central America, while CRM, for example, is but a second-fiddle career suitable only for those who are not good enough for "real" archaeology'.

Although many chapters are of interest to all archaeologists, it is easy to conclude that a student assigned one of these textbooks might think that 'real' archaeology is an exotic excavation in the jungles of Central America, while CRM, for example, is but a second-fiddle career suitable only for those who are not good enough for 'real' archaeology [4]. Furthermore, it is important to note that archaeology texts are much more than introductions for future archaeologists. The textbook, like the prospectus of an archaeology department, is an important interface between archaeology and the rest of the world (or at least that part of the educated world that may have taken a single archaeology class during higher education) and thus carries the responsibility/privilege of representing the field (see Tilley 1993b).

It is important, however, to ask how these texts are read. Although the author of the text might highlight certain issues at the expense of others, the author is 'dead', as the literary critic Roland Barthes commented (1974): the critical student may embark on an unintended reading or even toss the text into the rubbish bin, regardless of the author's meaning or desires. Does students' understanding of archaeology reflect the topical biases of the textbooks? There is indeed evidence that professors succeed, either through textbooks or lectures, in using their privileged position as teachers to propagate a view that favours their own interests as academic archaeologists, a view that elevates the archaeology of the academy and of the museum above all other archaeologies. Students prefer academic employment more than any other sector represented in the SAA, and research interests of students match most closely the interests of their professors and least closely the interests of public-sector and private-sector archaeologists (Zeder 1997a: 138-141). The results of the SAA survey reveal that archaeologists of all types prefer employment in a graduate university over any other job (Zeder 1997a: 60-61). Although this might simply mean that the types of activities conducted by academic archaeologists are the most preferable, other evidence corroborates the idea that academic archaeology is the most prestigious. For example, while most academic archaeologists felt that their current career path matched their expectations, public- and private-sector archaeologists (outside of academia) were more likely to feel that their current employment was inconsistent with what they expected to do (Zeder 1997a: 114), which shows that when archaeologists complete their degrees, they feel that academic archaeology is the 'proper' archaeological employment. In sum, academic archaeologists succeed in convincing many of their students that the archaeology of the academy is the most valuable type of archaeology. Perhaps the preference for jobs in a graduate institution as opposed to a two-year or four-year university signifies that other archaeologists are aware of the special discursive power that comes from being in a position to train those who will carry the discourse in the future -- today's postgraduate students.

Hiring practices and institutional prestige
Within academic archaeology, postgraduate programmes compete for prestige. Beginning with the fetishised assumption that capital breeds capital (Bourdieu 1988: 91), I now examine strategies that elite postgraduate programmes use to reproduce their academic capital. One strategy is inbreeding -- an elite institution hiring graduates of other recognised elite institutions. Armed with a list of the 12 most highly ranked graduate programmes in the US, based on a 1992 SAA survey that appeared in the 1993 SAA Bulletin (Appendix 2), and lists of the archaeology faculties of these 12 institutions, as presented in the 1997 Association of American Anthropology (AAA) Guide to Departments, I quantified the degree of elite 'inbreeding' by computing the percentages of professors, from each department, both full-time and part-time who received doctorates from the 12 most highly ranked programmes (Appendix 3) [5]. The highest degree of elite inbreeding among elite departments was at the University of California, at its Berkeley and Los Angeles campuses, where 90 percent of the professors received their degrees from elite institutions. The average inbreeding percentage for the top 12 was 75 percent (91 out of 121). In other words, 75 percent of the staff at elite departments in the US got their PhDs from elite departments. This figure is significantly greater than the percentage of elite PhD recipients employed as full-time or part-time professors in all universities listed in the Guide, which is 54 percent (480 out of 884) [6], and much greater than the percentage of total PhDs awarded by these top 12 programmes from 1968-1996, which is 39 percent (849 out of 2148). In sum, the percentage of people from elite graduate departments hired by elite programmes is much higher than the percentage of people from elite departments as a whole. It may be true that elite programmes hire products of other elite programmes because these 'products' are the best on the job market, but if that were the case, then why are not products of elite programmes hired in the same proportion across the board? My interpretation of this discrepancy is that elites hire other elites disproportionately to help boost their reputation.

The fact that elite departments prefer to hire PhDs from elite departments suggests that where one obtained one's degree is an extremely important factor in hiring practices in general. Obviously, there is a huge constraint on who gets to teach archaeology these days; in almost all cases, only those with a PhD can teach at the university level or above (Zeder 1997a: 55). However, it might also be the case that graduates of the most prestigious programmes are most often hired to teach. To evaluate this, I took a closer look at hiring practices between 1991 and 1996 (the most recent years for which data were available) and compared the number of PhDs awarded in this range of years to the number of these PhD recipients who held jobs listed in the 1996-97 AAA Guide to Anthropology Departments. These data were coded by postgraduate institution so that PhDs from the 12 'elite institutions' could be separated from PhDs from all other institutions. If there are constraints on which archaeologists are allowed to become professors and shape the discourse, we would expect to find that most of the teaching jobs available are held by people who receive their PhD from elite institutions. The data are presented in the following table. 'Teaching jobs' consist of either full-time or part-time positions, not includng adjunct positions.

	No. of PhDs awarded, 1991-96	No. of recent PhDs who found teaching jobs	Percentage of recent PhDs who found teaching jobs
Elite institutions	176	36	20.5
Non-elite institutions	326	44	13.5
TOTAL	502	80	15.9

As the table shows, people who received degrees from elite institutions have a better chance of finding a teaching job. If there were no 'institutional effect' on employment, we would expect that the percentage of people from non-elite schools who find jobs to be the same as the percentage of people from elite schools who find jobs. In fact, the deviation of the data from these expectations is significant at the 0.10 level of statistical significance (chi-square = 3.4711, df=1).

If we expand the range of analysis to include not just recent PhDs, but all PhDs from 1967 (the first year in which the AAA Guide to Departments listed this information) to the present, the trend becomes clearer. The relevant data are displayed in the table below. In this table, the '12 most successful schools' are not the same twelve 'elite' schools selected by the 1992 SAA survey and presented in Appendix 2, but those schools whose graduates between 1967 and 1996 held the most jobs as documented by the 1996-97 AAA Guide to departments. Not coincidentally, these two lists of 12 schools are about the same. The 12 'most successful' schools are listed in Appendix 4. The database of PhDs awarded begins with 1967 and continues to the spring of 1996, with a gap for the years 1969- 1970.

	PhDs awarded, 1967-96	Percentage of total PhDs	Jobs held	Percentage of total jobs	Percent chance of finding a job
12 most successful schools	849	39.5	369	49.3	43.5
All other schools	1299	60.5	380	50.7	29.3
TOTAL	2148	100	749	100	34.9

The information in the table again demonstrates that getting a degree from one of the top 12 schools improves one's chances of finding a job. If the school from which one received one's degree was not an important factor in finding employment, then we would expect those who received PhDs from the 12 elite programmes and from all other programmes to be hired in equal numbers (relative, of course, to the total number of PhDs awarded in each of these categories). However, the numbers presented above deviate from this expectation, and the deviation is highly statistically significant (p=0.001, chi-square = 31.5329, df=1). I underscore the fact that the most successful schools awarded 39.5 percent of the PhDs but took 49.3 percent of the jobs. Although all other schools awarded about 60 percent of the PhDs, these archaeologists hold only half of the teaching jobs.

Considering all job positions in the 1996-97 school year (not just those held by PhDs from 1967 to the present), we find again that some schools have many more of their PhDs placed in jobs. Appendix 5 shows the 12 schools that have been most successful in finding teaching appointments for their PhDs. (The list differs from the list in Appendix 4, because the Appendix 4 does not include teachers who received PhDs before 1967). When combined, these 12 schools control a majority of the archaeology teaching positions (full- or part-time) in the United States. The table below summarises this information.

	Number of PhD recipients holding teachings jobs	Percentage of total teaching jobs	Percentage of schools
Most successful schools (n=12)	480	54	14.6
All other schools (n=70)	404	46	85.4
All schools (n=82)	884	100	100

As the chart makes clear, 14.6 percent of the nation's archaeology graduate programmes control 54 percent of the teaching jobs -- a dominant and constraining presence in archaeological discourse. Of the 82 schools that have awarded PhD degrees between 1967 and 1996, seven programmes have been completely unsuccessful in finding teaching jobs for their degree recipients.

I mentioned above that the power of professors lies in their ability as gatekeepers to influence the archaeologists of the future. Because professors control (at least to some degree) the production of new archaeologists, whatever schools can control the reproduction of the professorial body will accrue the most academic power (Bourdieu 1988: 78). The disproportionate ability of certain graduate institutions to make professors out of their PhD recipients is therefore a source of academic power for those insitutions, and we should expect them to attempt to reproduce this advantage from year to year. However, the structure of rewards is never static; it must constantly be produced and reproduced (Giddens 1979). Competition for scarce awards ensures that the structure will be dynamic; the structure of inequality is never perfectly reproduced. In 1968, the first year the AAA published the doctoral affiliation of professors, 23 US graduate programmes succeeded in placing one or more of their PhD recipients in the teaching jobs listed in the AAA Guide. In 1981, 66 graduate programmes succeeded in placing one or more of their PhD recipients in teaching jobs. In 1996, 75 graduate programmes succeeded in placing one or more of their PhD recipients in teaching jobs. Despite this opening of the discourse to many more schools, in each of the three years examined here (1968, 1981, and 1996), the 'distribution of wealth' among these programmes was considerably lopsided. In 1968, 13.3 percent of the successful programmes held 54 percent of the jobs; in 1981, 12.1 percent of the successful programmes held 52 percent of the jobs; and in 1996, 14.6 percent of the total held 54 percent of the jobs. It thus seems that the job positions tend to cluster at the top. Although more and more programmes enter the discourse over the years, in each year a similarly small proportion of programmes controls an inversely proportional number of jobs.

In sum, there is a pattern of unevenness in hiring practices that constrains certain archaeologists (those from outside of the traditional top programmes) from participating in the discourse through teaching. We might interpret this unevenness by stating that the dominant archaeology schools have better professors and resources and therefore attract the best students, who are then most successful in finding jobs. There are many other elements that affect hiring practices that are not so easily documented, such as campaigning by professors, subjective preferences of the schools that hire, and any number of other political factors. However, I believe that the data presented here are strong enough to demonstrate that power structures and attempts to reproduce them are part of the many factors contributing to this unevenness, and that this discursive formation constraining to the benefit of those schools that are most successful. Although many archaeologists might deny that the reputation of their school put them where they are, being attached to a prestigious institution still seems to be a powerful resource for those seeking jobs. This is simply the power of a name. Therefore, I reach a conclusion similar to that reached by Bair, et al. (1986: 412) in their 1986 analysis of top-ranked general anthropology programmes: 'no more in [archaeology] than in the rest of the world do the deserving get their just reward'. As strategies of reproducing the power structure intersect with hiring practices, where one obtained one's degree can be more important on the job market than what one did to get that degree.

Citation pratices and rhetoric
Individual archaeologists also impose restrictions to gain prestige. Citation practices are an excellent way to create a bottleneck around a concept. By constantly citing oneself or failing to cite others, an author jockeys for ownership of ideas and becomes another kind of gatekeeper, assuring that future discourse must be related to the author's contribution (Becher 1989: 59-60; Foucault 1981: 64). As for citation practices, Tilley (1990b) observed that Michael Schiffer attempted to establish himself as a 'big man' in the archaeological tribe by citing himself 66 times (about once every paragraph) in a paper about the structure of archaeological theory. To this egregious example, I would like to add a recent paper (Spencer 1997), also about 'archaeological theory', in which the author cites himself 72 times. Non-processualists are equally guilty of what could appear to be self-promotional citation practices. For example, Hodder's initial critiques of processual archaeology (Hodder 1982, 1985, 1986) carried arguments similar to those published by Bayard (1969) and Kushner (1970), but he did not cite these prominent papers.

The debate over processual archaeology is also a good place to examine how rhetorical strategies constrain discourse. According to Foucault (1981: 64), 'Doctrine binds individuals to certain types of enunciation'. There is a reductionist tendency at work (Bourdieu 1988: 14), in which key words serve as academic capital whether or not they even retain any meaning (Baker, et al., 1990: 1; Moran and Hides 1990: 212). These words constrain creativity, because archaeologists have to place emphasis on these rather empty words as a strategy to conform to the requirements of the system of rewards. An excellent example is the word 'science' (see Tilley 1990a: 139-140). 'Scientific' New Archaeology was explicitly crafted as a reaction against the non-scientific archaeology that preceded it. (Of course, oversimplification, like what I have done in this statement, is a rhetorical strategy in itself.) In reviewing the rhetoric of the first issues of American Antiquity, I noticed, to my surprise, that many authors were also talking about a scientific archaeology in a polemical fashion (Ford 1938: 260; Mason 1938: 300; McKern 1935: 82; Steward and Setzler 1938: 10) [7]. The scientific archaeology of the 1930s was certainly different from the scientific archaeology of today, but in the theoretical debates of the 1970s and 1980s, New Archaeologists used the word 'science' without reflection on its complexity and historic meaning in archaeology. In this case, rhetorical strategies -- 'doctrine' in Foucault's terminology -- suppress the interesting and dynamic genealogy of the word 'science' and make it a buzzword used simply to gain currency. Another empty buzzword, this one used generously by processual archaeologists criticising the post-processual archaeology of the 1980s, is 'relativism' (Johnson 1998).

Another strategy I have noticed in polemical writings of various archaeologists is that of declaring archaeology 'difficult'. For example, George Cowgill writes (1993: 560), 'We must acknowledge the unavoidable difficulties, roll up our sleeves, as it were, and try to deal with them'; Richard Watson writes (1990: 673), 'The social sciences are truly the hard sciences'; Mike Shanks and Chris Tilley (1992: 107) -- 'Archaeology, and history, are the most difficult of all disciplines ...'; Robert Kelley (1992: 255) -- 'It would be nice if archaeology were easy, but it is not'; Bruce Trigger (1991: 552) -- '[Developing a holistic archaeology] is perhaps the most challenging and potentially important task ...'; Julian Thomas and Chris Tilley (1992: 107) -- 'the understanding of past societies on the basis of their material remains is surely one of the most complex philosophical problems human beings have ever set themselves'. What is funny here is that nobody in this recent debate has said that archaeology would be easy. This rhetoric of archaeology as difficult is therefore gratuitous and, as you can see, ridiculously exaggerated, but archaeologists have stuck to this rhetoric because whoever says that archaeology is hard can empower her or his voice by sounding sophisticated.

Coda
To conclude, I have modelled archaeology as a discursive practice and reviewed some of the ways power has shaped the discourse through constraints on knowledge production. I have attempted to emphasise, as does Bourdieu (1988), the strategic manner in which actors, motivated by socio-politico-economic interests, struggle to reproduce the discourse. Nevertheless, in cases like the discursive formation opposing women, there is no conspiracy of powerful actors dominating the discourse -- no smoke-filled room of archaeological bosses cooking up master plans for consent. Despite the fact that there are powerful actors that may achieve the status of a gatekeeper, in some cases it seems more likely that constraints to the discourse are perpetuated by subtle micro-processes diffused throughout the sites of practice. Parezo and Bender (1994: 74) seem to refer to these micro-processes when they state that 'subtle techniques ... take over as the major mechanism for limiting effective and coequal participation'. Equally subtle techniques seem to be at work in the teaching of archaeology. Professors may not even be discursively conscious of the topical biases of their textbooks. Likewise, the attitude of elite graduate departments toward hiring PhDs from other elite departments is probably neither overt nor recognised by those who perpetuate this practice. However, neither knowledgeable actors nor subtle micro-processes can fully dictate the order of discourse. Established structures may guide the discourse but do not dominate it entirely.

I suspect, however, that many readers are well aware of the conclusions that I have reached. Why, then, have I gone to pains to add extra details to a story that is already familiar? Some argue that this type of research can expose the historically constituted nature of truth and make it a target ultimately for a political critique (Baker, et al. 1990; Tilley 1990a: 327-8). In a similar vein, I consider this research useful for equity issues. To combat the chilly climate towards women in archaeology, for example, we must first make a detailed investigation of this climate to discover how and why equity is not realised. Beyond this, however, it would be useful to suggest ways in which the system could or should be reformed. True reform, I believe, is not possible. Since archaeology is a social profession, there can be no way of disentangling personal interests, institutional interests, and disciplinary interests from the pasts we interpret and produce. What we can do, however, is attempt to make these interests as explicit as possible. Although there is nothing new about this approach -- archaeologists have adopted an ironic trope (see White 1978) of critical self-reflexivity ever since the loss of innocence -- there are various ways to expose power structures in archaeological texts. One way is to recognise that archaeological texts, like all texts, are by nature intersubjective and intertextual. Every paper is driven by multiple motives and voices. Although many voices are acknowledged through citation, some motives are left unexplored. The reader is often left unaware of why and how a paper is getting published and the power relations embedded in textual production. Rather than suppressing these motives, we should call attention to them. This amounts to a sort of deconstruction that I will attempt to demonstrate by using my own paper as an example. A good place to begin is with the author. Who am I? The editors have asked me to append a 50-word biography to my submission. Such a biography should be part of the paper itself. Frankly, I am a postgraduate student facing a very difficult job market. I write papers like this and attempt to get them published partially out of concern for our field, but also in conformity to the system of rewards -- publish or perish. I have fretted over this paper to a great degree, knowing that it may offend 'the establishment', whatever precisely that is. Then again, do people who get offended by such papers even read an Internet journal like assemblage? Did this possibility guide my decision to submit this paper to assemblage? To continue the biography, I have recently entered the anthropology department at the University of California at Berkeley, a programme with a good reputation and success on the job market. Although I came here for many reasons, I am highly aware that my move might maximise my chances on the job market. Parts of this paper are written to please the reviewers of assemblage, specifically, this last section on moving beyond documenting the 'situation' to considering how to reform it. Other parts of the paper have been written specifically to conform to ideas that carry heavy academic capital in current archaeology. Specifically, I am referring to the many times I have cited French intellectuals like Foucault, Barthes, and Bourdieu and the somewhat unnecessary mention in this same paragraph of Hayden White's 'tropics of discourse', a topic which I take to be quite prestigious in theoretical archaeology these days, given its prominent role in the introductions to two recent volumes edited by archaeological theorists of the highest profile (Shanks and Hodder 1995; Tilley 1993a).

In sum, though strategies for prestige in archaeology may always mediate archaeological discourse, they can be understood if we try to make them transparent. Making these strategies and power relations transparent, perhaps through the sort of deconstruction attempted above, should be a top priority. Robert Preucel has recently suggested that archaeology can only begin its engagement with other social sciences and society at large if we reconstruct archaeology (conceptually) as a form of social practice (1995: 162-3). Along the same lines, Tilley has suggested that interpreting the past requires 'an understanding of the conventions and operations of that institution and discipline' that is archaeology (1993a: 15). To do this, we must recognise that what is said about the past, how it is said, and who says it, are always already mitigated by the power relations underlying the system of rewards in the present. Therefore, we must 'roll up our sleeves' and add to what we say about the past a critical study of the institutions, funding agencies, professional vehicles, etc. which on the one hand enable, yet on the other hand constrain archaeological practice.

Acknowledgements
I would like to thank Arthur Joyce for commenting on an earlier version of this paper and Fred McGhee for support.

1986 In Search of the Past: An Introduction to Archaeology (J. Bower).

1989 Archaeology, 2nd ed. (D.H. Thomas).

1993 Archaeology: Discovering Our Past, 2nd ed. (R.J. Sharer and W. Ashmore).

1993 Out of the Past: An Introduction to Archaeology (D.L. Webster, S.T. Evans, and W.T. Sanders).

1993 Archaeology: The Science of Once and Future Things (B. Hayden).

1995 Archaeology: An Introduction, 3rd ed. (K. Greene).