This paper was first presented at BMCR’s 30th anniversary celebration in October 2022.
I published my first review in BMCR in 1994 (1994.01.05), and I joined the editorial board in 1995. Along the way, I have published (so far) 36 reviews in BMCR, including at least one review in 23 of the 29 years since 1994. I have no idea how many reviews by other scholars I have edited for BMCR; there are 143 on my computer, but that is only since 2005. I have spent quite a lot of time thinking about the work of other scholars for BMCR, and (for the purposes of this talk) I ask myself why. Pure altruism it is not; and some of the people whose work I have reviewed would probably agree.
When I started reviewing for BMCR, three things particularly appealed to me. One was speed: if everyone was prompt, then a book could receive a review within months of its publication, and—more important to the reviewer—the review itself could appear within weeks of submission. Print is satisfying to look at, but it is never rapid, and I have had articles take so long to appear that I scarcely remember them. A second was the flexibility of BMCR about length: in 1998 I published a review of nearly 5000 words (1998.11.32), and in 2002 one of 5300 words (2002.02.09). The fact of not being constrained by paper, printing, and postage costs made BMCR generous about word count. The third appealing feature was ephemerality: the idea of writing something that had no physical existence was liberating; I felt less constrained in what I could say, because while it might be read soon by some people online, it would then vanish forever. BMCR is still pretty fast, although it no longer offers unlimited space, but I was wrong about ephemerality: some of what I have written in print is pretty hard to find, while BMCR is eternally and easily accessible.
In some respects I regret the loss of ephemerality: it led me to review books on subjects I knew relatively little about, in order to learn more; and it led me to write in a less formal style than a more permanent medium would have expected. Whether that was fair to the books I reviewed is a different matter; on the other hand, after having edited a journal myself, I have few scruples about sounding more authoritative than I have any right to sound. I have always found a useful maxim a couplet by Hilaire Belloc that (borrowing from Arnaldo Momigliano) I once quoted in a review: “O let us never, never doubt / what nobody is sure about.” Lack of knowledge is perhaps a hindrance, but it does not require great expertise in a subject to recognize a weak argument; and reviews that look at arguments are generally much more interesting and useful than those that simply look at content.
I have spent a good part of my career studying our distant predecessors, the Roman grammarians, and the four elements of the grammarian’s task still correspond pretty well to what we do as teachers, and even more closely to what we do as reviewers. The parts of grammar, as reported by the fifth-century grammarian Diomedes from Varro (who got them from some Greek grammarian), are lectio, enarratio, emendatio, and iudicium. In their original context, they describe the work of the ancient classroom: students learn to read and explain a text; they correct mistakes (in the original order, proofreading probably came before explanation, but Diomedes’ order works better here); and, finally, they offer opinions about its merits and defects, iudicium (krisis poematon, for you hellenists). That is what a reviewer also does: we read books carefully (lectio); and in our reviews we explain what the book is about (enarratio); point out the things that are wrong with it (and what is right, too) (emendatio); and finally offer some assessment of its merits (iudicium).
That is an idealized picture of what happens in the classroom, and even more idealized for book reviews. Anyone who has looked at ancient commentaries and ancient literary criticism will know all too well that they are long on enarratio but very short on iudicium, and the same is increasingly true of reviews. Far too many of the recent reviews that I have seen, and not just in BMCR, bear an unhappy resemblance to the formulaic five-paragraph essay American students are taught to write in high school. After an opening paragraph, we get enarratio, a chapter-by-chapter description of what the book says; and at the end, we get a brief concluding paragraph saying that this is a good book, makes a useful contribution, or even just summarizing the book again. I see many such reviews; I do not read them, unless the book they are reviewing is directly related to my own research. They are boring, and they contribute very little beyond what the author of the book probably says in the introduction or can be found in the preview on a publisher’s webpage. I am, of course, being unfair: a great many reviews actually do offer iudicium, but they seem to me to represent an ever-diminishing proportion of reviews, and the reluctance to commit iudicium, actually to say something that might include criticism, has spread its shadow far beyond the reviews published in BMCR.
Why write reviews? I can, of course, only speak of my own experience, but after having reviewed more than a hundred books, I feel entitled to pontificate. Motives change and vary depending on many factors. Let me begin with the practical. When I started reviewing, I had published very little, and I was flattered to be asked—and it gave me a publication, albeit a minor one, and that did not hurt. Nor did it hurt, at a time when my salary was not high and I had no research funds, to get for free a book I needed to read anyway. Over time, reviewing has brought me several thousand dollars worth of books, and that is not nothing; my prize, in that respect, was writing a 1200-word review of Bruce Frier’s magnificent edition of Justinian’s Codex, which lists for $725.00. For the same reason, I was once happy to read book manuscripts for publishers and get a few hundred dollars of books in exchange for a considerable amount of work.
From the crass to the sanctimonious: as I said a moment ago, I not only review books but, at least in the past, refereed a fair number of book manuscripts; I still referee perhaps half a dozen articles for journals every year; I have served on grant evaluation committees and promotion committees and written reports for tenure reviews. And I am not alone in any of this, nor is it surprising. It is simply an extension of the everyday exercises of iudicium we are all paid for: to make responsible judgments about our students and colleagues through grading, writing letters of recommendation, or hiring and promotion. This kind of writing is truly ephemeral (as well as often confidential), but it is not at all unimportant. I have written lots of letters, given lots of grades, interviewed lots of candidates—some of whom I remember, and for some of whom I hope I did some good. It is worth repeating this simple truth: what is expected of us every day, as teachers and as scholars, is the constant and careful exercise of honest iudicium.
Does reviewing do any good, and for whom? That is perhaps an awkward question. A favorable review may do the author of the book some good; an unfavorable one may have the opposite effect. But neither way, I suspect, makes much difference unless the review itself is worth reading: the five-paragraph essay variety of review does very little good for anyone, because it doesn’t say enough to help (or hurt, for that matter) the author of the book and it doesn’t say it well enough to do much for the reviewer, other than adding a line to a CV.
Then why write reviews? I don’t mean the short notices for Classical World or similar periodicals, limited to 500 or 600 words: those can include praise or blame, but they don’t have enough space for a reasoned argument, and all they can do is draw attention to a book’s existence. That is informative, but has little content. Longer reviews can actually say something, and not just summarize or criticize. Book reviews are the closest thing to essays that we get to write: they allow us to offer slightly speculative opinions, to use a book as a peg on which to hang thoughts on broader themes, to write prose without footnotes. If a book is good, we can use the review to place it in, or against, current opinion about a subject and to make suggestions about how it offers an opening for further developments. If a book is not so good, or has distinct weaknesses, one can offer alternatives—or even make it look better than it is by demonstrating its capacity to provoke ideas. A review is, ideally, a place to play with the author’s ideas—not in the way that a cat plays with a mouse, but in order to move the subject forward. Scholarship, ideally, is a collaborative enterprise, not in the sense of being done by committee, as far too much scholarship now is, but in the sense that one person’s ideas or discoveries almost inevitably start from something someone else has said. And a review can provoke that.
Perhaps my sense of what a review is or should be is out of date; we live in cautious and nervous times, and for fear of being themselves criticized, persecuted in social media, or sued, a great many people are unwilling to express honest opinions, particularly negative ones, in public. So too, a lot of scholars try to avoid the time-consuming thought and revisions that readers’ reports for journals often demand by publishing their work in collective volumes or conference proceedings, genres that have, to my mind, seriously degraded the quality of scholarship over the last generation or two (and I should say that I too have taken advantage of this): the articles in collective volumes are rarely carefully refereed, nor are they carefully reviewed. How often is a collective volume rejected because a small proportion of the articles it includes would on their own be turned down by journals? How can one give a serious assessment to ten or twelve articles in a review limited to two thousand words? The collective volume allows authors to avoid criticism, publishers to avoid doing serious editing, and reviewers to avoid the work of assessing a book’s argument. There are many and honorable exceptions to this, of course, but the desire to avoid saying anything critical, to avoid both judging and being judged, seems to me all too common.
So much for jeremiad; I will turn back to autobiography. I don’t remember which of my teachers it was, but someone, when I was an undergraduate, gave me three related pieces of valuable advice—valuable in the days before online databases, at any rate. One was that when I read an article in a periodical volume, I should look at (not necessarily read) all the other articles, to find out what the field was about and what questions interested people. A second was that, when I found an article or book I admired, to read more work of the same scholar, whether or not it was on a topic that concerned me, because good scholarship is worth studying for what makes it good scholarship; and along with that came the recommendation to buy, or read, Kleine Schriften or collected papers of scholars I admired. And the third was, when I read a book on a subject new to me, to read reviews of it.
I followed all three pieces of advice; and that led me to some instructive reviews and reviewers. One was a negative example: A. E. Housman’s reviews of practically anything. They contain great amounts of learning coated with glistening venom; they are entertaining, but generally more destructive than constructive. Unfortunately also very tempting, particularly to the young. A second was Eduard Fraenkel’s devastating review of Volume 2 of the so-called Harvard Servius, a paper I have read many times as an example both of how to review a critical edition, mixing details and broader observations, and of the way a reviewer can go beyond the book under review and say something about the subject that the book failed to say or (in the case of an edition) do. It is also a cautionary tale about the effect of a powerful review, on which I offer two anecdotes, which I have, I admit, used before. One is that it was nineteen years before the next volume of the Harvard Servius appeared—a volume that incidentally showed that the editors had learned almost nothing from Fraenkel’s review other than the need to collate manuscripts accurately—and the Blackwell’s catalogue, then the best English source for information about current publishing, added a snotty note to the effect that “we trust that nineteen years will not elapse before the appearance of the next volume.” They were wrong; it took more than fifty years. The other is that, when Wendell Clausen and I agreed in the 1970s to collaborate in completing the edition he had begun of the scholia to Persius, he said that it was safe to do it now that Fraenkel was dead. Some reviews terrify as much as they inform.
The third reviewer, whom I am lucky enough to have known and whose seminar I attended for a year, was Arnaldo Momigliano, and it is above all he whom I have tried, quite unsuccessfully, to emulate as a reviewer. I mention here just three reviews, that of Ronald Syme’s Roman Revolution published in the Journal of Roman Studies in 1940, and the paired reviews of Benjamin Farrington’s Science and Politics in the Ancient World and A. N. Sherwin-White’s Roman Citizenship, published next to one another in the next JRS, in 1941. These are astonishing reviews, filled with learning in multiple languages and fields, making remarkably careful judgments about the books he was reviewing, critical and respectful at once—and, I add, published when Momigliano was in his early 30s, had lost his Italian chair and been in England, without a fixed academic position, only since 1938. But what I would emphasize here is not Momigliano’s great learning and wisdom, but that all these reviews, while recognizing the genuine importance of the books under review, make suggestions about topics about which he believed the author had not said enough: the review of Sherwin-White begins by praising Roman Citizenship and explaining its place in the literature of Staatsrecht; it then proceeds to list no fewer than twelve areas that deserve further exploration. The reviews of Farrington and Syme are similar, if not written in the form of a list, and they are still quoted in more recent scholarship for the acuteness of Momigliano’s judgments and suggestions.
What a good review does is to show why a book is right or wrong or innovative or suggestive, why it matters—or, when appropriate, why it fails to matter; and in doing so it not only helps readers understand the book, it explains a field of study in a way that can draw non-experts in, and make them more expert. Even if reviews like this are critical, they are also protreptic. I do not have Momigliano’s range and ability as a reviewer (or as a scholar, for that matter); it is the protreptic element of his reviews I admire as much as the scholarship, and to read them makes me want to know more about the subject. I know colleagues who are reluctant to take the time to write reviews and who are reluctant to be critical when they do write them, but that is a mistake. To show that we take a book and its subject seriously and to show why we take it seriously is what scholarship is about—and even more important, it is what teaching is about.
 Hilaire Belloc, “The Microbe” in More Beasts (for Worse Children) (London, 1897) 48; it is the last couplet in the last poem of the book. It was cited by A. Momigliano, “Cavalry and Patriciate: An Answer to Professor A. Alföldi,” Historia 18 (1969) 388; from which I cited it in CW 109 (2016) 438 (review of D. Feeney, Beyond Greek: The Beginnings of Latin Literature [Cambridge MA, 2016]). In the oral version of this paper I wrongly attributed the lines to Ogden Nash; I am grateful to James O’Donnell for the correction.
 “Grammaticae officia, ut adserit Varro, constant in partibus quattuor, lectione enarratione emendatione iudicio . . .” (Varro, fr. 236 Funaioli = Diomedes 426.21K). The proper order in Quintilian (1.4.3) and several other grammarians, but without Varro’s name. For the meaning of these officia, cf. J. Zetzel, Critics, Compilers, and Commentators (New York, 2018) 18-20.
 M. C., “Witty Chissels,” TLS (July 2, 2021) 27 collects several recent complaints about reviewing, and criticism of reviewing (generally of non-academic reviewing) has been going on for a very long time. For a bracing attack, see Elizabeth Hardwick, “The Decline of Book Reviewing,” Harper’s Magazine (October, 1959) 138-43; there are valuable discussions of reviewing also in M. Reich-Ranicki, Kritik als Beruf : Drei Gespräche, ein kritisches Intermezzo und ein Porträt (Frankfurt, 2002). I am grateful to Cosima Mattner for the last two references.
 CW 111 (2017) 154-56.
 I mention with appreciation some reviews that have corrected and improved on my own work: M. D. Reeve, CP 70 (1985) 85-92; J. Grant, Phoenix 39 (1985) 86-88; R. J. Tarrant, CQ 58 (2008) 480-83; U. Schlegelmilch, BMCR 2006.01.24, as well as the extended criticism of S. Timpanaro, Per la storia della filologia virgiliana antica (Rome, 1986). I would rather not have needed correction, but to provide a starting point for other scholars’ thoughts is not a bad thing.
 I cite two outstanding reviews (which I read in the week before I delivered this paper) that show the possibility of transcending the general weakness of reviews of collective volumes: John Ma’s review of M. Gygax and A. Zuiderhoek, eds., Benefactors and the polis: the public gift in the Greek cities from the Homeric world to late antiquity (Cambridge, 2020) in BMCR 2022.10.36, and S. Douglas Olson’s review of C. Stray, M. Clarke, and J. Katz, eds., Liddell and Scott: The History, Methodology, and Languages of the World’s Leading Lexicon of Ancient Greek (Oxford, 2019) in Gnomon 94 (2022) 596-601.
 My own experience is that editors of journals are happy to receive careful assessments from referees whether they are positive or negative, while editors of books tend to sound very unhappy to receive negative assessments. That may mean no more than that managing the assessment of a book takes more time and money than managing the assessment of an individual article.
 A selection of some of the more vigorously nasty passages of Housman’s reviews and prefaces may be found in John Carter, ed., A. E. Housman: Selected Prose (Cambridge, 1961); all the reviews are included in J. Diggle and F. Goodyear, eds., The Classical Papers of A. E. Housman (Cambridge, 1972)—a collection made in contravention of Housman’s own wishes, as expressed in his will (and quoted by the editors).
 Originally published in JRS 38 (1948) 131-43 and 39 (1949) 145-54; reprinted in E. Fraenkel, Kleine Beiträge zur klassischen Philologie (Rome, 1964) 2: 339-90.
 Review of Syme, JRS 30 (1940) 75-80; review of Farrington, JRS 31 (1941) 149-57; review of Sherwin-White, JRS 31 (1941) 158-165. All three reprinted in A. Momigliano, Secondo contributo alla storia degli studi classici (Rome, 1960), respectively 407-16, 375-88, and 389-400.