Although Rocke's study is an outstanding work, there are three or four themes in his work with which I disagree, and these will be the focus of this critique.

Although Forbidden Friendships was published in 1996, it is based upon Rocke's doctoral dissertation presented to the State University of New York at Binghamton in 1990. In other words, it was written in the late 1980s during the height of the fashion for the theory of social constructionism, at a university which was a centre for the modern sociological approach to homosexual issues. I think it is because of this historical context (i.e. the Ph.D. student being overly sensitive to the latest theories in his field of research, especially coming from his own university) that Rocke presents standard social constructionist "conclusions" that do not really arise from his evidence.

Rocke's book is based upon a thorough study of the records of the Office of the Night (Ufficiali di notte) which was active in Florence from 1432 to 1502, which contains documents about more than 16,000 men implicated in sodomy, of whom nearly 3,000 were convicted. (He supplements this with other material as well.) Thus, in a city of only 40,000 inhabitants, the majority of men were incriminated for engaging in homosexual relations at least once during their lifetimes. This is an unparalleled wealth of documentation, virtually unexplored until Rocke investigated it, and Rocke handles it with expert skill (and with the aid of computer analysis). None of my remarks should be understood as challenging his data; my argument is with the way he has interpreted it, or, more precisely, the way he has refused to adopt a common-sense interpretation of it.

Rocke begins to go badly wrong at the outset, when he sets this material in the context of modern theory. He makes the standard assertion that "People of the Middle Ages and early modern period lacked the words to convey the precise equivalents of the current 'homosexuality' as a distinct category of erotic experience or 'homosexual' as a person or a sexual identity." (p.11) And yet he immediately acknowledges that "The terms 'sodomy' and `sodomite,' which were standard in the religious and juridical language of premodern Europe for conveying same-sex relations, might however seem to work as substitutes, for in some contexts they appear to have much the same meanings.' (p.11, my emphasis) And he points out that Thomas Aquinas defined sodomy as comprehending all sex between persons of the same gender (p.11) — this of course is equivalent to the modern abstract definition of homosexuality. But then he insists that "outside the subtle field of moral theology, however, 'sodomy' was not strictly synonymous with 'homosexuality,' nor was 'sodomite' the equivalent of the noun 'homosexual'." In support of this claim, he points out that Florentine laws against sodomy applied to women as well as men, and that "it was certain sexual acts alone that denoted sodomy, not (as in prevailing theological views) the gender of the persons who practiced them" so therefore "a sodomite was not, strictly speaking, a person who engaged in sex with members of his (or her) own biological gender; the sodomite, that is, was not a homosexual, but a person who committed the various acts defined as sodomy." (p 12) But then he immediately proceeds to contradict and undermine his whole argument: "This generally being the case, when Florentines used the words 'sodomy' and 'sodomite' in a generic way they probably had sexual relations between males in mind, since these were by far the most common and conspicuous, and aroused the greatest public concern." (p.12). In other words, to all intents and purposes, a 'sodomite' was equivalent to a male homosexual. Rocke's statistical evidence further establishes beyond all possible doubt that Renaissance Florentines understood "sodomy" to mean homosexual relations between men irrespective of a specific sex act, and understood "sodomite" to mean "a man who has sex with men". In other words, after all the theoretical reservations that Rocke has put forth, we have ended up with the common-sense (and in my view correct) application of the modern concepts of homosexuality and the homosexual to this earlier period.

The records documented by Rocke suggest to me that, contrary to Rocke's own view, late medieval Italians organized their understanding and representation of homosexuality in much the same way that modern Europeans do today. Rocke's introductory claim is that late medieval Italians differed substantially from us in their understanding of homosexuality because of "the absence of conceptual categories based on sexual object choice" (p.12). Yet the records he produces make it very clear that sexual object choice was central to the experience of Florentine sodomites, and to people's understanding of them. This is proven in at least two ways: Firstly, although homosexual relations were markedly age-structured, the younger partner was post-pubescent, i.e. sodomites desired boys – that is, young males rather than children of indeterminate sex – who already exhibited the secondary sexual characteristics of males and were obviously male. Secondly, a statistically significant number of cases mention fellatio, i.e. the older partner sucked the penis of the younger partner before sodomizing him. The fact that a boy's penis was important in the relations contradicts the standard argument that boys and women are desired indiscriminately because both are treated as passive sex receptacles. It also indicates, among other things, that giving the boy pleasure was important to the active sodomite, i.e. he did not simply "take his pleasure", but it was a mutually pleasurable relation. Of course the laws were constructed in such a way that anal intercourse would be more highly documented than oral intercourse; since the law specifically regulated sodomy, it is significant that such a high incidence of oral intercourse is documented, and likely that the reported incidence would have been even greater if the law had been constructed to regulate oral intercourse more specifically.

Having asserted in the introduction that in Renaissance Florence sodomites were not conceived of as constituting a distinct sexual minority, and having asserted that the Florentines "had no way to distinguish verbally, either with these official designations or with other vernacular terms, a man who engaged in sex with males from one who committed precisely the same acts with women, whether occasionally or even exclusively" (p.12) – in other words, having asserted that they lacked conceptual and verbal categories for "the homosexual" – in the very first chapter of his book Rocke goes on to demonstrate just the opposite: they did in fact have so-called "reductive" labels for more or less fixed sexual categories and identities (and in later chapters he will inadvertently present the statistical evidence for these more or less fixed sexual categories and identities).

For example, from a very early period the law dealt with a "special class of offenders – men who engaged in sodomy with several partners or habitually over a long period. Often such a man was labeled a publicus et famosus sodomita" (p.23); Rocke cites a typical case in 1352 involving Miniato di Lapo, known as "a public and notorious sodomite" who pursued his relations "publicly and openly, with many, many boys" (p.23). Rocke throughout the book cites many cases of men who were conceptualized as a special category of sodomites who "regularly", "publicly and openly", and "notoriously" enjoyed sodomitical relations. Throughout the legal records we regularly come across the distinctive phrase "a public and notorious sodomite" (p.24), and in the more vernacular terminology of their accusers they were called "a dedicated sodomite" (p.25). Many such men had long affairs with specific youths (e.g. lasting two years or more) (pp.24-5). Rocke gives very good pen-portraits of several of these "notorious sodomites", such as Salvi di Niccolo Panuzzi, whose sodomitical career spanned nearly 30 years, and who at the age of 63 confessed that he had solicited young men to sodomize him (a reversal of the usual role) (p.79).

Rocke has to acknowledge that the sermons of Bernardino of Siena in fact contain "acute observations" (p. 36) rather than constructed stereotypes: "more often than not, the wealth of information from court records in the later part of the century corroborates [Bernardino's] descriptive remarks." The whole basis of Bernardino's sermons is that the sodomite was a very clearly recognized class of person. Bernardino argued that sodomites did not want children of their own, and claimed that they were the reason why the Florence birth-rate was lower than any other city in Italy, and had declined considerably in the past twenty-five years (pp. 36-7). Tuscany, and Florence in particular, was notorious throughout Italy for its sodomites, so much so that Tuscans would not be hired elsewhere as schoolteachers "for fear they would corrupt the boys" (p. 37). Bernardino felt that parents encouraged sodomy in their boys through their lack of moral guidance (and that some fathers even taught sodomy to their sons) (p. 37). Bernardino castigated mothers who did not look too closely into the source of their young sons' purse of money, because he knew they came from their sodomite admirers. (p. 38). Judicial records demonstrate that sodomites "typically bestowed gifts of money, clothes, or other items on the adolescents they courted or sodomized" (p.38). Bernardino critized parents for encouraging their sons to dress fashionably, in tight and flashy clothes: "It's a serious crime, to have a short doublet made for them and stockings with a tiny patch in front and another in back, so that they show a lot of flesh for the sodomites. . . . Send them out [dressed] decently, not like girls!" (p. 38). He criticized mothers for "effeminizing" their sons with fancy clothes and refined manners: "it appears you make your son look like yourself, so that to you he's quite becoming: 'Oh, isn't he the handsome lad!' and even 'Isn't he the pretty girl!'" (p. 38). Rocke points out that he has not found a single court record actually noting the existence of transvestism among boys (p. 38), but a type of "girlish" clothing was clearly recognized by Bernardino and felt to be characteristic the boys desired by sodomites. But the main point is that Bernardino clearly recognized two categories of sodomites: those who engaged in homosexual relations in their youth but gave it up when they grew older, and those who remained incorrible into their old age, whom he called "inveterate" sodomites (p. 39). These two categories based upon life stages are also clearly documented in the court records investigated by Rocke. But Rocke refrains from acknowledging that these are the two classic categories used in most homosexual sexological literature: youthful experimentation in which the boys are not really homosexuals, and adult sexual behaviour by men who really are homosexual, reserving the term "homosexual" more properly for the latter category. (Restif de la Bretonne in Les nuits de Paris (late 18th cent.) observed what has since become a commonplace concerning opportunist or situational homosexuality: "In this class must be included schoolboys who do it for mischief, soldiers for lack of money, and monks of necessity. As for mignons, it is certain that they do so only from avarice".) Bernardino's recognition that a man who did not give up sodomy beyond the age of 32 or 33 was an "inveterate sodomite" is not really much different from the modern sexological recognition that after that age such a man could be classified as a "congenital invert". Bernardino even recognized, as the early sexologists argued, that it was impossible to alter the desire of the adult sodomite: "The devil blinds him so that if he passes 33 years of age, it's nearly impossible for him to reform. He can, but it's very hard to stop . . . it's nearly impossible." (p. 40).

Rocke observes that "Florentine men normally put off marriage unti the average age of thirty or thirty-one, and a large proportion never took a wife" (pp. 28-9). This supposedly fostered illicit or "unauthorized sexual activity of all sorts" and sexual debauchery (p.29). Bernardino of Siena in his famous sermons of the mid-1420s "insisted that sodomy deterred young men from marrying, that bachelorhood fostered sodomy, and that most sodomites were unmarried" (p.29). "Woe to those men who fail to take a wife when they are the right age and have a legitimate reason! . . . For if they don't marry they become sodomites. Make this a general rule: when you see a grown man in good health who doesn't have a wife, you can take this as an evil sign about him" (p.40). This argument by Bernardino demonstrates a clear awareness of a heterosexual/homosexual binary. There is hardly any indication in Bernardino that a man can be bisexual: Bernardino almost always sees a clear boundary line between marriage and sodomy, i.e. between what we today would call a heterosexual and homosexual orientations. Contrary to the view that sodomites are simply men who occasionally engage in sodomitical acts, for Bernardino, sodomites were men with erotic dispositions, men who specifically "were not aroused or attracted . . . by even the most desirable of women". Contrary to the view that sodomites desired both boys and women, Rocke finally admits that "Bernardino believed that the sexual tastes of at least some sodomites focused one-dimensionally on males and more or less precluded relations with females" (p. 41).

Rocke's evidence gives statistical support for the link between bachelors and sodomites. In 1403 public brothels were actually promoted to prevent the excesses of sodomy (p.30), and in such circumstances (i.e. favourable to heterosexuality) it is not clear why illicit sexuality included so much sodomy. Rocke suggests that the low marriage rate led to sodomitical activity, but this seems to me to set the cart before the horse. Rocke makes it clear, for example, that the Florentine Commune did many things to make marriage attractive, such as creating a dowry fund to help less wealthy families marry off their daughters. (p. 29) And of course they also took many steps to discourage, suppress, and even eradicate sodomy. But the simple fact seems to be that young Florentine men were averse to marriage, and not averse to sodomy.

Tied in with his doctrinaire assertion that sodomites did not constitute a distinctive sexual minority, Rocke also argues that "the organization and character of homosexual behaviour . . . suggests that this sexual underground [i.e. of sodomites] did not constitute a separate world or a truly distinctive 'subculture'" (p.15). Rocke is adamant that the practice of sodomy in Renaissance Florence "probably did not resemble anything like the highly visible, organized subcultures of the modern world populated by a consciously distinct and coherent category of persons who today might be called 'homosexuals,' an anachronistic model that hardly applies to these traditional societies" (pp. 26-7). I find this claim really difficult to understand in the light of abundant evidence supplied by Rocke about institutions, behaviours, and characters that one would normally associate with a subculture. For example, "buco" (hole) is a common pun for the arsehole, and a Florentine tavern called the Buco was run by a man named Antonio Guardi who was depicted in a satirical poem ("La buca di Montemorello" [monte is a metaphor for buttocks, and buco, hole, is a metaphor for the arsehole] in which this real person was depicted as a sodomite. There was "judiciary evidence shows that in the later fifteenth century, the Buco was a common haunt of sodomites" (p. 33).

In fact, Rocke repeatedly establishes evidence that "networks among sodomites . . . were common and extensive", for example in developing political client relationships (p.57). Rocke apparently rejects the word "subculture" because it is an imprecise term, but his term "network" is hardly more precise. More importantly, he fails to appreciate that numerous historians and sociologists have found it very useful to use the word "subculture" in its broadest meaning, to suggest, for example, ethnic enclaves or people on the "margins" of society: other synonyms whose precision need not be insisted upon are "underworld" and "demi-monde". Rocke seems to think that a genuine subculture has to be entirely separate from the mainstream culture, but very few sociologists ever use the word in this strict way. Words such as "demi-monde" are very often applied to sexual subcultures, notably prostitution rings or networks, and Rocke provides abundant evidence of such homosexual networks of prostitution: for example, a circle of more than 50 men were discovered to have had sexual relations with a single 16-year-old lad named Andrea (p.68): he obviously went the rounds of sodomitical cruising areas, and probably many of these men passed him along to their fellow sodomites. It is certainly true that when a sodomite learned that a lad he had sodomized was giving evidence, the sodomite alerted his fellow sodomites so that they could confess themselves (and thereby escape a heavy penalty): "It was probably no coincidence, for example, that the same day the Night Officers received an anonymous charge in 1482 named sixteen-year-old Baldovino di Giovanni Baldovini, six men showed up to denounce themselves for having sodomized him" (p.70). There was obviously a gay grapevine that worked very quickly and efficiently. Rocke is not at all consistent in his rejection of the concept of the sexual subculture. For example, he seems to have no misgivings about using the word "haunt", which is commonly associated with sexual subcultures, and he frequently identifies many "well-known haunts of Florentine sodomites" such as the street of the Furriers (p.78).

Rocke's dual contention is that male homosexual relations were fully integrated into Renaissance Florentine life, and that their understanding of homosexual relations was wholly different from our modern understanding: "sex between males was a common and integral feature of daily life in this city, [and] it formed part of a universe of experience and values that differed substantially from our own." (p.10) Men were not separated into homosexuals and heterosexuals (p.10). They did not share the modern "notion that a person's homosexuality or heterosexuality profoundly defines one's personaltiy and identity" (p.11). "Italians of the Renaissance would have found current beliefs about homosexuality and heterosexuality as well as much of modern sexual experience foreign indeed" (p.11). In other words, Rocke claims that there is an epistemological boundary between premodern and modern.

It is probably quite correct to conclude from the evidence that the officials "neither pursued nor punished sodomites consistently or rigorously. Their restraint probably reflects a fair degree of popular and official accommodation, if not outright tolerance, of sodomy in Florence" (p.73). However, it is a very large jump to the further view that "Homosexuality was deeply integrated into that cluster of social structures, gender values, and forms of aggregation that together helped constitute male culture in Florence" (p.15). Rocke gives abundant evidence that sodomites were regarded as being separate from their society, and that their society took means to forcibly separate them from society. For example, in 1496 some 50 parishioners chased their sodomitical priest from the church (p.83); this is just a typical incident of many cases in which local people took the law into their own hands. Neignbours regularly joined in a chorus of ridicule of shopkeepers whose sodomitical inclinations were known to them (p.83). There were many instances of sodomites being chased down the streets by people and boys calling "Burn them! Burn them!" (p.203) – which hardly supports the claim that homosexuality was part of male culture as a whole. The Florentine women also frequently complained that their husbands were engaged in sodomitical relations on the side – which again contradicts any claim that sodomy was simply an accepted feature of sex in general.

The social constructionist notion that premodern Italians did not recognize the heterosexual/homosexual binary seems rather easy to disprove. For example, Antonio Beccadelli in 1425 wrote his two-part Hermaphroditus, part I of which celebrates love between males, and part II of which celebrates love between males and females (p. 43). Not only is this a characteristic example of the hetero/homo formula, but the fact that Beccadelli was consequently accused of being a sodomite and forced to burn his book also shows that society was hardly "indifferent" to the two forms of love, but violently condemned the homosexual part of the binary. And with regard to subcultural geography, Rocke shows that the taverns most noted for sodomitical activity were not noted for female prostitution (p. 161) – in other words, there is very little polymorphous or indiscriminate licentiousness, but two forms of licentiousness, one specifically homosexual and the other specifically heterosexual.