Corps de l’article

Le point de départ […] doit être un retour aux évidences biologiques : l’espèce humaine est mâle et femelle. Sur ce donné naturel se sont construites des règles sociales […] l’opposition homme/femme sert de modèle, d’archétype à toutes les oppositions abstraites (froid/chaud, dedans/dehors, etc.)

Françoise Dekeuwer-Défossez[1]

Les mots qui […] désignent, par exemple les noms de couleurs, le jour et la nuit, l’homme et la femme, construisent des objets qui n’existent pas dans la nature avec la netteté que les concepts fabriqués par l’homme y impriment. À quel moment passe-t-on du jour à la nuit ? Même les mots fabriqués pour qualifier les situations intermédiaires, l’aube, le petit jour, la tombée du jour, le crépuscule, sont plutôt des images de poètes que la délimitation de situations précises. La distinction des sexes n’échappe pas à pareille indétermination.

François Rigaux[2]

Many of us take for granted that the human species is divided — that we are divided — into two neat groups : male and female. This ostensible fact is reinforced by the sale of blue or pink (but never both) baby clothes, by an array of expected behaviours and social roles that match the blue and pink clothes — and by the inevitable Ms or Fs that appear on our driver’s licences, passports, and other official documents. The pages that follow focus on this third aspect of sexual differentiation : the state’s (permanent) classification of its citizens as (only) male or female.

But what does it really mean to be male or female ? The modern answer is that sex is determined by a combination of factors. Some are “objective” : chromosomal sex (XX or XY), gonadal sex (testes or ovaries), external morphologic sex (penis and scrotum or vagina and breasts), internal morphologic sex (prostate or uterus), and hormonal patterns (predominantly testosterone-based or predominantly oestrogen-based). Others are “subjective” : self-identified sex (psychologically male or psychologically female), performed sex (engaging in male-type acts or female-type acts) and relational sex (treated as male or treated as female[3]). Harmony between all of these elements is both assumed by and constitutive of official designation as male or female[4]. Yet this notion of harmony belies the incredible variety both within and among the factors mentioned above[5].

Two particularly well-known instantiations of such variety are transsexuals and intersex persons. Transsexuals are difficult to capture with a single definition[6], but as a working definition we can assert that transsexuals subjectively identify with a sex other than that which their objective characteristics would suggest. Intersex persons, meanwhile, are born with objective characteristics not easily classed as exclusively male or female : for example, their genitalia may be simultaneously or between male and female in appearance or function.

The law, however, has ignored these and other examples of disharmony within the factors that reflect or determine sex, choosing instead to class all citizens as strictly male or strictly female[7]. Doing so has required the state to make choices about which factors are determinative of sex[8] — yet these choices are rarely explicit and, as I will illustrate below, often contradictory. Nonetheless, “the law presumes a binary sex model[9]”.

Legal privilege may be given to the male/female dichotomy because “[w]estern culture is deeply committed to the idea that there are only two sexes[10]”. As Johanne E. Foster reminds us, “sex […] classifications have been used to determine who could vote, who could own property, and who could legally be denied access to certain occupations and educational institutions. Today, one’s access to employment, education, political power, and even leisure time free from the demands of housework and childcare are all still directly related to one’s sex[11]”. However, tradition is an insufficient reason to continue a practice (think capital punishment), and not all tradition needs to be reflected in the law (think “Honour thy father and thy mother”).

This paper thus contests the state’s classification of its citizens by sex[12]. Though many have argued that sexual classification yields unfairness to various groups, very few people have advocated the abandonment of sexual classification[13] ; most of those who do address such a proposal brush it off as no more than an ephemeral possibility[14]. Only Dean Spade has actively promoted the desertion of permanent sexual classifications[15], but he does so in the American, common-law context. To date, no one has examined the value (or lack thereof) of permanent sexual classification in Canada or in the civil law. I have thus chosen Québec as a doubly valuable locus of investigation.

I begin in Part 1 by describing the cognitive importance of categories, pointing out the perniciousness of applying categories to people. Part 2 outlines the importance and the effects of categorization by the state, as well as the prevalence of sex as a classifier. Part 3 zeroes in on Québec, describing how the province and its civil law system consider sex as an important and nearly immutable component of each citizen’s civil status. In Part 4, I zoom back out to examine how governments like Québec’s might justify their choice of sex as a classifier. I consider science-, identification-, socialisation-, and intervention-based justifications, concluding that none holds water. Part 5 outlines possible responses to this failure and advocates the dissolution of permanent sex classifications. In Part 6, I describe the effects of such dissolution. I conclude that ceasing to mark each citizen indelibly as male or female would have no detrimental effect on the Québec government’s ability to craft policy and could instead foster more nuanced, more accurate, and fairer state intervention.

1 Thinking Through Categories

Geoffrey C. Bowker and Susan Leigh Star begin their book on Classification and Its Consequences with the simple assertion that “[t]o classify is human[16]”. Assuming that they are correct, however, Bowker and Star’s statement prompts a triumvirate of key questions : “What do we classify ?”, “How do we do so ?”, and “Why ?” Addressing these three general questions serves as a necessary first step in this specific enquiry into the sexual classification of the citizens of Québec[17].

At its simplest, a category is “a spatial, temporal, or spatio-temporal segmentation of the world[18]”, involving “judgments to the effect that one or more objects possess, or lack, one or more characteristics[19]”. Mary Douglas and David Hull expand on this notion of similarity, writing that classification is a process of discarding “the heterogeneous material […] from one emergent specialised domain after another and ma[king it] irrelevant by classificatory decision[20]”. Such a process may be vitally important : “Without the ability to […] categorize, we could not function at all, either in the physical world or in our social or intellectual lives[21]”. In other words, it seems that we classify all objects according to their similarities — and that we have no choice in so doing.

Though categorization may be central to thought, however, it is the necessary dependence of categorization upon assessments of similarity that may render the process of categorization detrimental to lucid thought. Similarity is “relative and variable, as undependable as indispensable” because

when to the statement that two things are similar we add a specification of the property they have in common, we […] remove an ambiguity ; but rather than supplementing our initial statement, we render it superfluous. For […] to say that two things are similar in having a specified property in common is to say nothing more than that they have that property in common. [Therefore,] comparative judgments of similarity often require not merely selection of relevant properties but a weighting of their relative importance, and variation in both relevance and importance can be rapid and enormous[22].

For example, determining that from amongst a gooseberry, a strawberry, and a piece of shortcake, the gooseberry and strawberry are “most similar” involves an assessment (and thereafter an assumption) that reproductive function and edibility — reflected in the shared “berry” name — are more cogent as classificatory criteria than is the manner in which something is eaten.

No matter that the criteria linking gooseberries and strawberries are supposedly scientific or objective : “Classification is usually treated as an outcome of an ordering process as if the organization of thoughts comes first, and a more or less fixed classification follows as its outcome. But the ordering process is itself embedded in prior and subsequent social action. It is a middle part of a circle of questions and answers[23].” The “classical theory” of categorization, developed by Aristotle and Kant[24] and reliant on “[c]lear [b]oundaries, [s]hared [p]roperties, [u]niformity, [i]nflexibility, and [i]nternal [d]efinition[25]”, has thus been replaced by an experiential model grounded in the “naturalization” of categories[26]. Over time, categories become “entrenched[27]”, rendering them doubly “invisible[28]” : not only do we rarely recognise that we are categorizing, but when we do, we assume that the categories are natural and appropriate[29]. Thus, the categorization of reality — or, perhaps, of the objects that populate our respective realities — depends on a value-laden ordering of otherwise neutral components or characteristics[30].

Such ordering applies not only to fruits and food but to people — and Lakoff warns us that categorization of people “can be pernicious[31]”. Pernicious, it is argued, for three reasons : “[m]ost categorization is automatic and unconscious[32]” ; “any theory of categories presupposes the possibility of an ontological difference between a thing and its Being[33]” ; and categorizing people inevitably creates a “social and moral order[34]”. In other words, classifying persons is an invisible process by which their characteristics are abstracted from themselves and used to assemble them into groups in a social hierarchy — and the boundaries of these groups are defined by the person or persons doing the categorizing, not by the persons within the categories.

The perniciousness of categorizing people becomes especially pro- blematic when one sees classifications as not only (artificially) descriptive but as prescriptive : “the system’s description of reality becomes true[35]” because categories constitute the characters of their constituents. The “moral and aesthetic choices” embodied by categories “craft people’s identities, aspirations, and dignity[36]” in that “[b]y using classification evidence is weighed, ambiguities are discounted or eliminated, claims are adjudicated[37]”. From declaring that “Sam is an X-type person” (to use a relatively value-neutral example, a Canadian person), it is but one small step to assert that “Sam is X” (Canadian), thus erasing both the elements that make Sam different from other (people who are) Xs and the elements that make Sam (and other [people who are] Xs) similar to (people who are) not-Xs. This elision of “form, matter, and individual substance[38]”, however, “leaves out” imagination in favour of reduction[39], and denies Sam — and others — full agency in their self-determination. Categories are further entrenched as the material consequences of categorization are incorporated into and strengthen the definitions of the categories themselves[40] : as all Xs are all treated in a certain manner (as all Canadians are given Canadian citizenship documents), it is a fortiori expected or even demanded that all Xs subscribe to such treatment to “count” as Xs (being Canadian requires citizenship documents).

The “naturalization” of categories is only questioned by “problematic cases[41]” or “boundary objects[42]” that exist alternately or simultaneously in multiple categories. Where such objects submit to the classificatory scheme, the questions are silenced ; where, however, “an object refuses to be naturalized,” the result is a “monster” that upsets the “laws of nature” by espousing “madness or heresy[43]”. Monsters implicitly challenge the very value or necessity of categories, “rais[ing] worries about the foundations of knowledge[44]”. The only escape from these worries, perhaps, is community-wide subscription to an “ethics of ambiguity[45]”, whereby categories are permitted to develop and to evanesce, and to have fuzzy boundaries in the meanwhile. Such a result, however, seems unlikely not only because of classification’s importance to thought but also given the persistent state interest in establishing and maintaining categories of persons.

2 Sex and the State

James C. Scott provides a compelling account of the symbiotic relationship between categorization and the growth of the modern nation-state. Pre-modern states, he writes, knew little about their subjects, and as such their interventions either involving or directed at these subjects were “often crude and self-defeating[46]”. As time went by, however, “officials took exceptionally complex, illegible, and local social practices […] and created a standard grid whereby [intra-state activity] could be centrally recorded and monitored[47]”. This grid included last names, maps, cadastral lists, and standard units of measurement, all of which “reduce[d] an infinite array of detail to a set of categories that [would] facilitate summary descriptions, comparisons, and aggregation[48]”. The result was increased “legibility”, by which Scott and his colleagues refer to the “capacity to locate citizens uniquely and unambiguously” and “standardized information that will allow [the state] to create aggregate statistics[49]”.

This legibility was, however, not just descriptive but functional : as suggested in Part 1 above, the categories chosen by the state to facilitate legibility “represented only that slice of [society] that interested the official observer” and, combined with state power, “would enable much of the reality they depicted to be remade[50]”. Popular “resistance” was overcome by “[t]he increasing weight of the state in people’s lives and the state’s capacity to insist on its rules and its terms[51]”. Thus could the state validate its classifications through their normalization : as state interventions became not only accepted (and even desirable) but commonplace, the classifications on which these interventions were based became perceived as legitimate.

Lest we think that state classification is entirely malicious, Paul Starr reminds us that :

Every regime needs to draw lines between kinds of people and types of events when it formulates its criminal and civil law, levies taxes, and allocates benefits. In contemporary politics, classification is particularly important in economic regulation, the collection of social statistics, decisions about legal standing in class-action suits, and the design of insurance rates, pensions, and selection criteria for jobs and university admissions[52].

Classifications, often “mediated through mundane bureaucratic documents such as forms[53]” and “modulated by local administrative procedures[54]”, are constructed by governments — and here I borrow Spade’s term — to do some kind of work[55], and the work of government is to shape society. This element of “political choice” thus distinguishes governmental categories from other classificatory schemes[56].

For a government’s classification scheme to do good work and to do it well, however, not only the classifications themselves, but also the use to which they are put, must be perceived as legitimate : they must be generated and maintained by principles that citizens believe (or become convinced) are reasonable[57]. This concern with legitimacy underpins s. 10 of Québec’s Charter of Human Rights and Freedoms, which enumerates prohibited grounds of discrimination — that is, illegitimate uses of classification[58]. The Charter’s s. 86, however, shields from illegitimacy any “affirmative action program [designed] to remedy the situation of persons belonging to groups discriminated against” in employment, education, or health services. A successful remedy requires that individuals be identified as members of such groups. The result is “classificatory tension[59]” between efforts to reject some (uses of) categories and efforts to maintain others ; in other words, the legitimacy of the categories and the work done with and by them is called into question.

But what is the effect of such efforts — and such crises of legitimacy — on individual citizens ? The connection between a person and a category may be “chosen or imposed”, and may not be within an individual’s power to change[60]. This incapacity is especially true of categories that are “tightly coupled with a person” — and one such category is sex[61].

In the section that follows, I will address the tight coupling of legal sex designation to each individual within Québec’s civil law tradition. Before I do so, however, it is worthwhile to recap the theoretical ground covered thus far :

  1. Classification is central to thought, but the choice of categories is both motivated by and constitutive of social reality ;

  2. As social rather than scientific facts, categories are subjective and involve moral choices ;

  3. Unlike individuals, governments can enforce participation in their subjective classificatory schemes through state power ;

  4. Modern governments create and maintain categories for the purposes of doing work, and the legitimacy of this work may be contested ;

  5. As part of this work, individuals may be tightly bound to categories, one of which is sex.

3 Québec, Classification, and the Code

The government of Québec, like the governments described generally above, has a need for legibility amongst its citizens : “Comme toute personne est appelée à jouer un rôle sur la scène juridique, elle doit être identifiable[62].” However, unlike other Canadian governments, that of Québec operates within a civil law framework — a framework that places particular emphasis on the stability of law. Jean-Étienne-Marie Portalis, the primary drafter of the Napoleonic Code (ancestor of Québec’s laws), wrote that “au lieu de changer les lois, il est presque toujours plus utile de présenter aux citoyens de nouveaux motifs de les aimer[63]”, thus tying stability to legitimacy. The very act of publishing an ostensibly comprehensive code establishing the jus commune that underpins all other laws[64] suggests a normative stability in contrast with common-law notions of legal evolution and statutory exceptionality.

In the civilian system, legal stability is achieved by using a civil code to divide the normative world into categories into which all aspects of the noumenal world are placed. “In such an approach, organizational choices about where a particular topic should be elaborated and decisions about the relative prominence afforded to any topic are as much a part of the Code’s interpretational logic as is the manner in which individual rules are actually formulated[65].” That Book One of the Civil Code of Québec addresses “Persons” must therefore be understood to place the law of persons in a position of primacy within Québec’s normative structure, in terms of not only importance but also logical progression : Book One “erects a protogenic conception of the physical person as a titulary of legal rights that is largely independent of any other definitions[66]”.

Within Book One, after instituting the enjoyment of civil rights and the existence of certain personality rights related to bodily integrity and privacy, the C.C.Q. moves quickly to establish and to regulate the “Status of Persons”. In the French context, Jean-Paul Branlard defines a person’s civil status as “l’image juridique de la personne […] il évoque l’idée de répartition, de classement, en assignant à chacun sa place dans la société[67]”. Already, we see the tight linkage between civil status and classification. Whether couched in terms of identity control[68] or public order[69], the goal of tracking civil status is unquestionably related to legibility : the state has an interest in identifying (key aspects of) its citizens in order to effect political interventions[70]. By extension, therefore, the state has an interest in the stability of civil status[71], as fluctuations in status could lead to illegibility and ineffective intervention[72] — and ultimately to illegitimacy. Branlard describes civil status as in principle immutable (that is, incapable of change), “indisponible” (that is, “[l]e titulaire de l’état ne peut à son seul gré le modifier”), and imprescriptible (that is, the law will not accommodate prolonged behaviour at odds with civil status[73]).

In Québec, an individual’s civil status has four components — name, sex, family status (single, married, or in a civil union), and residence — of which the first three are attested and registered by acts of civil status[74]. Of the four components, however, sex is often ignored : some doctrinal writers do not even mention sex in their discussion of civil status[75]. This omission likely stems from how “obvious” it seems as a classifier : “Le sexe fait incontestablement partie de l’état des personnes [et] constitue le premier critère de répartition statistique de la population[76]”. Sex is “le partage primordial” because, unlike the other components of civil status, which must be researched, assigned, or justified, “chaque être humain […] porte sur lui, dans la vie quotidienne, qu’il est un homme ou une femme[77]”.

Once used as a classifier, sex is unquestionably binary : “Toute personne, même si elle présente des anomalies organiques, est obligatoirement rattachée, à la naissance, à l’un des deux sexes[78]”. Every individual, legally, must be either male or female — and “en droit, l’hermaphrodite n’existe pas[79]”.

However, though it assumes that sex is binary, the C.C.Q. (like Québec’s other laws) nowhere defines how sex is to be determined or recognized[80]. The result is an inconsistent assessment of various objective and subjective factors, with objective factors predominating in determining legal sex at birth. Once determined, legal sex is difficult if not painful to change[81], as understood by many amongst Québec’s transsexual and intersex populations.

For transsexuals and intersex people are the “monsters” described in Part 1 above. They upset extant categories and shake the foundations of knowledge — here, the government’s knowledge of its citizens and the legitimacy of its classifications. And though the vocabulary of “monsters” may seem unfair, it is not unlike that used in doctrinal treatises of civil law : transsexuals suffer from a “maladie[82]”, they are “diabolique” and “un danger pour l’espèce toute entière[83]”.

What is the appropriate response to this threat to the stability and legitimacy of categories ? The traditional responses are either to blur the boundaries of the existing dichotomy or to create additional categories between “male” and “female”. Another possibility, raised much less frequently but in favour of which I will argue in Part 5 below, is to abandon sexual classification altogether. Before any solution can be proposed or espoused, however, we must examine the basis of the existing classification. If all state-based classification is designed to do work, then what work does sexual classification do ?

4 Why Sex ?

4.1 Sex Reflects Science

Simply put, it could be that sexual classification reflects the scientific truth that humans are differentiated by genetics, anatomy, and hormones into two groups, and that the behaviour of these two groups is correspondingly different. On several levels, however, the claim of scientific difference is problematic. First, the assertion that humans and other animals exist in dimorphic (that is, two forms : male and female) distinction is disproven by the facts. On a chromosomal level, “[s]exual differentiation in mammals requires a precise choreography of molecular and cellular events” ; the choreographer is evolution, in that male-associated traits enable males to survive better while disadvantaging females, and vice versa[84]. Not infrequently, however, the choreography skews from its expected trajectory and yields diversity within dimorphism[85], resulting in not only XY (male) and XX (female) chromosomal pairings but X (female), XX (male), XYY (male), XXY (male), and many others[86]. In terms of anatomy, which is largely dictated by genetics, “[w]e are born into a physiological continuum on which there is no discrete and definite point that you can call ‘male’ and no discrete and definite point that you can call ‘female’[87]”. As to hormones, a survey of vertebrate sexuality shows that “no fundamental uniformity exists that is regularly apparent” between male and female animals, such that “we must be willing to entertain the possibility that some dimorphisms have neither a genetic nor a hormonal basis[88]”.

Turning to the potential effect of dimorphism on the actions of males and females, “[few] behaviors [other than], e.g., the ejaculatory pattern, are present in one sex but not in the other[89]”, so “[w]hat seems a female behavior in some species is a male behaviour in others[90]”. Having particular sex characteristics “does not insure normal sexual behavior[91]”, where “normal” is understood to signify concordance with expected behaviour for a given sex. Thus, even if “there are clear, reproducible mean differences in many neuroanatomical variables when groups of male and female brains are compared[92]” (and here note that the differences are “mean” and not “absolute”)[93], we do not know how and whether these differences translate into behavioural practice[94].

The result of an analysis of these ostensible differences is that, in some ways, males and females are different because of their biology ; in others, they reach similar biological or behavioural outcomes by different paths ; and in still others, they are different in ways that are beyond the reach of hormones or sex-specific genes[95]. Albert Jacquard thus reminds us that “[t]he chief lesson to be learned from genetics is that the groups to which we belong do indeed differ from each other, but that the individuals within each of these groups are even more different still[96]”.

Additionally, the expectations of scientists may colour their interpretation of any experiment yielding supposedly sex-based differences. Any effort to “solv[e] […] non-normative behavior […] shapes the design and interpretation of experiments[97]”, just as these experiments inherently involve a preconception of what is “normative”. Thus, “appeals to what is biologically natural are deeply embedded in our cultural beliefs about the meaning of […] sexuality[98]”, some scientists even go so far as to superimpose male/female binaries onto organisms in which differentiation has nothing to do with sexual function[99].

It seems, then, that there is no firm scientific basis to distinguish all males from all females, and that attempts to do so may be disingenuous. Even if there were scientific differences between men and women, however, where is the benefit in permanently declaring all children “male” or “female” at birth so as to represent these differences legally ? It is hard to believe that the work done by sexual classification is simply a record of biologically based mean outcomes. Like sex, ethnicity is genetically determined and arguably “permanent” — yet Québec and other governments decline to include ethnicity on any government documentation. Like sex, mental illness can yield behavioural differences — but mental illness does not appear in any standardized government data bank in Québec. Sexual classification, then, must do some other work.

4.2 Sex and Recordkeeping

Key to the immutability of sex in the civil law and the legibility of the Québec population is the web of documentation on which sex is displayed and, therefore, the interlocking array of data banks in which sex is recorded by the government of Québec. These data banks are used for surveillance and registration purposes, for distribution of government services and benefits, and for population-wide statistical assessments — in other words, for purposes of legibility. Because of their standardised nature, government documents are also used widely in the private sector as forms of identification and sources of personal data. Every one of these documents, and by extension every one of the underlying data banks, includes information on legal sex :

  • Birth Registration and Birth Certificate. The sex of every person born in Québec is recorded by the accoucheur at the moment of birth and transmitted to the Directeur de l’état civil (DEC) through an attestation of birth[100]. The parent or parents also include the child’s sex in a declaration of birth submitted to the DEC[101]. Together, the attestation and declaration permit the DEC to register the civil status of the child, to issue documents (including a birth certificate) confirming this civil status, and to initiate the process of obtaining health insurance. The birth certificate, once issued, displays the child’s sex ;

  • Health Insurance Card. Cards are issued automatically by the Régie de l’assurance maladie du Québec (RAMQ) once the DEC transmits birth details to the RAMQ ; these cards display the child’s sex[102]. Those who are not born in Québec but want to obtain a card must register with RAMQ, indicating, inter alia, their sex ;

  • Driver’s Licence. Those who wish to obtain a Learner’s Licence require a birth certificate, passport, or proof of Canadian Citizenship or Permanent Resident Status, as well as a Health Insurance Card — all of which display legal sex[103]. Graduation to a Probationary Licence requires presentation of one of these pieces of identification[104]. A full Driver’s License is issued automatically on completion of the probationary period. All three types of licence display the bearer’s sex. Further, to acquire a Driver’s License Plus for cross-border travel to the United States, an applicant must : show a birth certificate or Certificate of Canadian Citizenship or Naturalization ; present a Driver’s Licence, Health Insurance Card, Certificate of Indian Status, or passport (all of which display the bearer’s sex) ; complete a Consent Form that requires disclosure of sex ; and agree to the transmission of personal data including sex to the American government, which keeps these data for 75 years[105].

In addition to these regimes maintained by the government of Québec, citizens of Québec also have their sex recorded and displayed by several regimes operated by the federal government :

  • Certificate of Canadian Citizenship. Acquiring a Certificate requires disclosure of sex on the Application Form, as well as presentation of a birth certificate, Driver’s Licence, or other equivalent[106] ;

  • Permanent Resident Card. Obtaining a Card requires presentation of a passport or travel document, as well as a Driver’s Licence or similar government-issued document[107] ;

  • Certificate of Indian Status. Applications must be supported by a birth certificate, as well as a passport or two other forms of identification such as a Driver’s Licence[108] ;

  • Passport. To obtain a passport, each applicant must provide a birth certificate or Certificate of Canadian Citizenship, plus one of the following : Driver’s License, Health Insurance Card, Certificate of Indian Status, government identity card, or existing passport. In addition, the passport Application Form requires disclosure of sex[109].

Unlike their American equivalents, Canadian Social Insurance Number (SIN) cards, and related data banks, do not display or contain information on the bearer’s sex[110]. The legislation that authorizes the collection of SIN data, however, permits them to be used “for all purposes for which a Social Insurance Number is required[111]”. This open-ended language has permitted the SIN to stray “beyond its intended purpose as a file number for government programs” such that it is now used for government purposes as varied as cashing savings bonds and establishing a Registered Education Savings Plan — and for private-sector purposes including credit checks, employee benefits calculation, movie rentals, and pizza delivery[112].

Of particular concern is “[t]he growing use of data matching, comparison and exchange between and among jurisdictions in administering social programs[113]”. This data matching, most often based on SIN data, has “implications for both privacy protection and program integrity” that have prompted considerable concern[114]. Amongst these concerns is not only the leakage or unintended use of data but the permanent linkage of one data bank to another.

In other words, though the SIN card may not include sex, SIN data are linked to enough other government data banks, and each of these banks is in turn linked to enough others, that every Québec citizen’s sex is not only part of but is linked to and accessible by a wide variety of government — and perhaps private — programs at many levels. All of these programs attach (or can attach) a binary sex classification to an individual citizen, and they all trace back to the individual’s sex as listed on the birth certificate[115].

What happens, then, if this citizen does not conform to the contours of the binary classification itself ? How does this web of personal information, collected and collated across jurisdictions and over time, respond to people who refuse to be slotted — or to slot themselves — into the tidy, exclusive categories of “male” and “female” ? Such people appear in considerable biological and psychological variety, of which two significant (but by no means the only) groups are transsexuals and intersex persons[116].

Traditionally in the civil law, the “indisponibilité” of civil status has prevented transsexuals from demanding that the government recognize their destination sex and reflect this sex on official documentation[117] : “L’état civil du transsexuel ne saurait être modifié puisque son sexe génétique ne peut jamais évoluer[118].” This insistence on stability, based on the primacy of chromosomal sex, persisted even in the face of international jurisprudence to the contrary[119], and could even extend to finding doctors who conducted sex reassignment surgery to be guilty of criminal mutilation[120]. More recently, however, civil law jurisdictions including Québec have accepted legal sex change : a Québec resident of majority age having undergone “surgical operations involving a structural modification of the sexual organs intended to change his secondary sexual characteristics” may have this change reflected in a modified birth certificate after applying to the DEC and providing two attestations from physicians[121]. The success of such “action[s] en réclamation d’état[122]” shows that the indisponibilité of civil status has weakened, and that civil status may be less than fully immutable[123].

In theory, an altered sex designation on the birth certificate should permit a person to apply to have this change reflected in all other documentation and data banks. The inevitable delays in this process across multiple agencies, however, can cause particular concern where data matching is prevalent : if the records of a person’s civil status in two separate data banks do not align, this person may be “flagged” for particular inspection or for denial of benefits or services[124]. Further, the process of changing legal sex may not be as simple as filling in a form and awaiting a response ; other bureaucratic and social hurdles may await[125]. More broadly, however, if sex is nowhere defined in the C.C.Q., but is understood to involve an array of objective and subjective factors, why should the focus on one particular objective factor (morphology of secondary sexual characteristics) necessarily be determinant[126] ? In other words, if courts and the government have abandoned chromosomal sex as decisive of legal sex, how and why is morphological sex any more reasonable an indicator ?

Intersex persons cloud the question further. As legal sex is determined at birth by means of objective criteria, intersex persons challenge the binary, exclusive sex categories of “male” and “female”. The traditional response to intersex babies has been to “fix” their genitalia surgically and then to assign a corresponding legal sex[127]. If the initial assessment or surgical choice is belied by later (that is, chromosomal) evidence, the person’s civil status can be altered — a process that philosophically involves not a true change but rather a correction : “mettre l’état civil en accord avec la réalité[128]”. Réalité is thus here defined as chromosomal correspondence with the XY (male) or XX (female) category, as “le sexe chromosomique est immuable[129]” and “[ne] laisse aucune marge de manoeuvre[130]”. This notion, however, is at odds not only with the truth of genetic variation but with the abandonment of the chromosomal determination of sex as applied to transsexuals.

We thus see that legibility, as reflected in government documentation and as recorded in data banks, rests on shaky foundations. There is little clarity or consistency as to which factors truly “count” in determining an individual’s sex for recordkeeping purposes. The result of the search for — and imposition of — legibility is thus more uncertainty rather than less. If legal sex is not therefore doing the work of identifying citizens, however, then perhaps it serves another purpose.

4.3 Sex as Social Engineering

Perhaps the government distinguishes between men and women in order to ensure the continuity of differentiated social roles[131]. By requiring that sex be displayed on official documentation, the state can ensure that citizens are constantly reminded of their identity, and of the behaviour expected of those who inhabit such identities. Others who look at this documentation — be they police officers pulling over an errant driver or potential employers recording details about a candidate[132] — may adjust their own comportment in response to the M or F that they (expect to) see, thus ensuring that sexually essentialized behaviour is socially supported and reinforced[133].

Given legislation against discrimination by sex[134], legal reforms designed to remove sexual distinctions[135], and scrupulously sex-neutral language[136], the government of Québec cannot be charged with explicitly promoting role differentiation by sex. If it does so implicitly, then such a strategy merits exposure and rejection in light of the principle of sexual equality.

Perhaps, on the other hand, Québec’s government identifies men and women for social purposes designed not to pigeonhole but rather to benefit both sexes. In the paragraphs that follow, I will discuss legal sex as it applies to public health, segregated facilities, and affirmative action[137] — each of which is ostensibly geared towards aggregate welfare.

4.4 Sex-Based Interventions

Turning first to public health, it cannot simply be that “people can realize their greatest potential for happiness and productivity only if they are sure they belong to one of only two acknowledged sexes[138]”. Any health-based motivation for sexual classification must therefore be broader in scope. Thomas Insel sees a public health-related value in studying sex differences[139], in that certain diseases, conditions, and other maladies may affect women and men differently. Acknowledging this variable susceptibility, however, does not require permanent attachment of individual citizens to a sexual classification : Spade points out that heart disease also affects susceptibility, but is considered an “individual aspect of medical history” rather than a permanent marker[140]. Further, “if a government program is interested in tracking uterine cancer rates, perhaps more accurate information will result from tracking the rates of this cancer in people with uteruses than in people who are socially classified as ‘female,’ since those two categories are not identically matched[141]”. An example of a more nuanced approach is a San Francisco health care system using “six categories for its classification of sex, depending on the patient’s genetic type, bodily type (which may be surgically altered) and presentation of self[142]”. These last two examples demonstrate that binary sex classifications may not best serve the interests of public health.

A second ostensibly beneficial intervention based on binary categorization is the creation and maintenance of segregated facilities. Arguments in favour of such facilities typically centre on comfort and safety[143]. Sex-segregated facilities such as bathrooms, however, may serve as targeted locations for violence, rather than as havens from violence[144]. Behaviour within sex-segregated institutions may be more violent and hierarchical than that in mixed institutions, and sex-integrated prisons may correspondingly be less violent[145]. Sex-segregation might not therefore make people safer, and it most certainly can make some groups (including homosexuals, transsexuals, and other gender-variant persons) significantly less safe and less comfortable[146]. Segregation based on dichotomous sexual distinctions may not thus do its supposed work — and, even if it did, one could reasonably ask whether permanent attachment of a given person to a given sex would be a necessary precursor to successful sexual segregation.

Finally, sex segregation may figure in affirmative action or other ameliorative programs designed to better the lot of long-disadvantaged groups including women[147]. Claiming membership in categories, however, tends to increase strife and to reify these categories to the disadvantage of others[148]. Thus, locking a person into an identity as “woman” and then allocating her a particular benefit may breed resentment both within and without the category of women[149] : men may feel excluded, but more importantly for this discussion the struggles of transsexuals (whether formerly or currently women) and of intersex people — as well as of gender-variant women, lesbians, and numerous other “sub-categories” that are both overlapping and unrecognized by the state — may be totally ignored[150].

If the work of classification is to enable targeted intervention, then permanent male/female categorization is likely far too blunt an instrument to get the work done well (if at all). A better approach would be to undertake data collection “with an understanding that what is being measured is the impact of social processes of gender production that result in discrimination and exclusion […] the gender categories used in such collection might not simply be ‘male’ and ‘female’ depending on the kind of problems being assessed[151]”. Interventions could then be structured that : considered classifications carefully in each situation to determine if they corrected past injustices or contributed to further injustice ; recognised unnecessarily categorical thinking ; attempted to address both similarities and differences within and between groups ; recognized the groups engaged in oppression ; and acknowledged that the social construction of categories does not render these categories meaningless[152]. Thus could better work be done.

5 Rethinking Sex

I have tried to show above that the permanent, binary classification of men and women by the government of Québec cannot accommodate the biological, behavioural, and psychological diversity of Québec’s population. At the same time, such indelible classification does little work — and what work it does could be done better. In other words, not only do the categories “male” and “female” not function well, but in their permanence they do not serve their function well[153].

A sensible response to this double dilemma would be to re-evaluate the categories themselves : if “[t]he explanatory value, and hence the status, of a paradigm is threatened by anomaly, […] before people abandon old paradigms, someone must articulate an alternative paradigm that accounts convincingly for the anomaly[154]”. In the paragraphs that follow, I thus consider four possibilities : clarifying the contents of the categories “male” and “female” ; widening the categories ; creating intermediate categories ; or abandoning permanent assignment to categories. I conclude that abandonment is preferable for a variety of reasons.

5.1 Sharpening Boundaries

If the Québec government is currently inconsistent in the criteria that it uses to undertake sexual classification, and if these criteria when applied result in imprecise or ineffective intervention, then one response could be to clarify the criteria while ensuring that they function with fairness. In other words, Québec could draw a bright line between “males” and “females,” and then could police this line vigilantly to secure justice.

To the bright line first : sharpening the sexual boundary would require that the government choose which indicators of sex should predominate from amongst those described above, and then regularize and publicize this choice across all domains of governmental classification and action. In some ways, this development would be retrograde : Roman law permitted intersex individuals to be classed according to their dominant physical characteristics[155], and early English law did the same[156]. In the modern context, however, choices would soon arise that would be difficult if not untenable : a return to the primacy of chromosomal sex would run counter to decades of progression towards full rights for transsexuals, while a focus on morphological sex would fail to account for the remarkable variation amongst human bodies. Both options, as well as others involving alternative objective indicators or combinations thereof, might still place people in compartments that “feel” or “seem” wrong to themselves or those around them. A more subjective approach, meanwhile, would involve a degree of choice (and, pending maturity to make such a choice, no small amount of uncertainty) at odds with the stability sought by the government in an individual’s civil status.

These deficiencies are unlikely to find a remedy in “fairer” application of the criteria. Fairness here could entail applying sexual categorization consistently to all citizens (that is, in maintaining the integrity of categories at all costs). Such an approach, however, not only would result in injustice to some individuals involved — one has only to imagine a male-to-female transsexual housed in an all-male prison, or a woman denied access to an affirmative-action program because she has an XY chromosomal pairing — but would run contrary to recent equality jurisprudence[157]. Alternatively, fairness could consist in plucking individuals from one category and depositing them in the other. In this scenario, however, the creation of exceptions to the rigid rule would entail bureaucratic or judicial calculi that should ideally have been incorporated into the initial act of classification. The existence of such exceptions to this ostensibly complete rule would thus undermine the very notion of a sharper sexual divide.

5.2 Blurring Boundaries

If a more rigid male/female dichotomy could be not only unjust but self-defeating, another possible response to the failure of sexual classification as currently undertaken by the Québec government would be to loosen the bounds of each category such that entry and exit were made easier. Under such an approach, individuals could be categorized by their dominant characteristics at any given time, with “characteristics” here understood to encompass all components of sex, objective and subjective. Michael S. Kimmel recognizes the varying importance of these components when he writes that, though primary sexual characteristics are more decisive at birth for sexual classification, secondary characteristics and behavioural representations are those that we observe in our daily interactions with others[158]. In other words, the proposed expansion and overlap of categories would elide the distinction between sex (“the biological apparatus”), gender (“the meanings that are attached to those differences within a culture[159]”) and performance[160], and would simultaneously give greater weight to the “concept philosophique de l’identité, i.e. [le] droit à l’autonomie du sujet[161]”.

Such an approach would not be without its challenges, however : New York City scrapped in 2006 a plan to allow legal sex change without proof of surgery, partly out of concern over the knock-on effects on segregated institutions and partly out of the need to conform with the federal Real ID Act[162]. Non-conformity between Québec’s legal sex requirements (or practices) and those of the other provinces or of the Canadian government could be equally problematic. Further, shifting the boundaries of “male” and “female” might do very little to address the disjuncture between the categories’ work and their effects as described above : making it easier for people to be male or female does not address the problems inherent in being only male or female at a particular moment.

5.3 Adding Boundaries Between

To address these concerns, some have suggested the addition of intermediate categories between “male” and “female”. Again, this concept is not new : the medieval De Spermate allowed and explained the intermediate category of hermaphrodites[163], and Randolph Trumbach asserts that the reductive two-sex model did not emerge until the early eighteenth century[164]. Modern writers including Julie A. Greenberg have espoused the creation of a third sex[165], while Anne Fausto-Sterling at one point proposed five[166]. Kimmel tells us that some other cultures already have three or four sexual categories[167].

Could such an expanded classification do its work better than the male/female dichotomy ? Perhaps, but likely not. Adding categories would be more, but not completely, reflective of the sexual continuum. It would not necessarily facilitate movement between categories — and could quite possibly do the opposite, if finer granularity is equated with greater stability. Targeted interventions could perhaps be more accurately constructed, but this is a weak argument for permanent attachment of individuals to categories if temporary categorization would suffice.

Creating additional categories could also have unintended consequences : it could further reify the existing boundaries of “male” and “female”[168] ; it could correspondingly reify the boundaries of the new middle ground[169] ; and it would still fail to account for the situations of those individuals in whom the components of sexual identity are not consistent with the criteria defining the categories. Further, there is a practical argument against expanding the number of categories : there may be too few people in the medial categories to justify creating particular facilities or institutions, and the result could be the functional collapse of the medial categories into the current male/female dichotomy. Creating additional sexes could lead to new forms of sexual discrimination[170]. And, finally, there would again be the issue of compatibility between Québec’s data banks and those of the other provinces and the federal government.

5.4 Removing Boundaries

A more radical and more reasonable solution might be to abandon permanent governmental sex classification entirely. As Spade points out, if the classification is not doing the work that we expect of it, it might be worth asking whether such classification is necessary at all[171]. This is not to say that “individual privacy rhetoric [should] valorize an end to government data collection[172]” ; such an extreme would be detrimental to the possibility of positive state intervention. Instead, we can imagine a state in which sex is recorded on a “need-to-know basis”[173]. Health care providers could ask a patient’s sex and record it in a file, much as they do with family medical history, allergies, and blood type. Segregated facilities could be re-examined and redesigned on the basis of efficacy and humane treatment rather than on the assumption that sex segregation is requisite. Affirmative action programs could be more nuanced so as to target subpopulations — sexual, visible, or other — rather than assuming, for example, that all women “need help” in a given situation.

The central claim here is that there is no need to attach a sexual marker indelibly to each person and to display this marker on documents meant to last a lifetime. Instead, we could “shift toward a more critical view of the use of gender data in government recordkeeping[174]”, a view that would not necessarily discard sexual classification entirely, but would accept temporariness, movement between categories, and the usefulness of defining and recording sex differently in different circumstances.

Sex could thus remain as a classifier where useful and appropriate. Individuals could indicate their sex on applications for government employment in order to facilitate workplace diversity initiatives[175], and injured persons could be attended by physicians with access to medical files including sex alongside allergies and medical history. No one need assume, however, that sex is requisite to all state intervention : we could imagine prostate cancer studies that target those with prostates rather than those labelled as “men”, unsegregated bathrooms in which everyone feels comfortable and safe, and identification cards that do not display to the world an M or F that is more embarrassing than the picture next to it.

6 Law Without Sex

But what would the effect of such a change be ? Would government data collection fall apart ; would the legal system as we know it collapse ? Likely not. Unlike a situation in which Québec collected sex data using different criteria or added a third sex category, the simple omission of a datum from each Quebecer’s entry in a data bank would not result in “translation” problems when data were compared across jurisdictions. Such discrepancies already exist : for example, British Columbia collects and displays information about individuals’ weight on their driver’s licences[176], while Prince Edward Island does not[177].

Even where laws are differentially applicable to men and women, the absence of a permanent classification should not be problematic — or could perhaps even prompt larger questions about the validity of differential applicability. The only way in which men and women are treated differently under the criminal law is in the infanticide provisions[178], which apply only to “female” persons despite the fact that legally male persons may give birth ; surely the drafters of the Criminal Code did not intend such an omission. In private law, recent amendments to the C.C.Q. have tried to make even the concepts of maternity and paternity as neutral as possible[179] ; the only other sex-based difference relates to surrogacy contracts, to which the same logic applies as with the criminal infanticide provisions just described. Where a will is made out to “my son John” or “my eldest daughter”, the principle of the liberal interpretation of wills in favour of intended beneficiaries should lead courts to override sex classification-based language[180]. Though men tend to receive higher tort damages, only men can apparently win claims for loss of consortium, and only women seem to succeed in actions for breach of promise to marry, judges in tort “normally accept persons as male or female as they appear to the courts or as they represent themselves”, so problems are unlikely to arise[181] — and we should in any case ask why tort law treats men and women differently in the first place[182].

More important than changes in individual laws or their consequences, perhaps, would be the effect that ceasing sex classification in favour of more nuanced approaches would have on society. If sex were no longer seen as an absolute classifier, it would follow other classifiers like race and class — which the state neither records nor considers fixed — in becoming one of many “categories of analysis” rather than an inherent feature of all persons[183]. For the state to cease seeing those currently labelled “men” and “women” as intrinsically different would be a significant step in the progression towards sexual equality — and also a result of this progression — even if it by no means would solve all sex-related social problems[184]. Ceasing permanent sexual classification would permit disclosure of sexual identity where an individual saw fit and was comfortable to do so, rather than demanding disclosure and often humiliating the individual.

I do not mean through this argument to call for the abolition of sexual identity, nor do I believe that discarding permanent sex classification would leave citizens identity-less and thus rob the state of its ability to intervene amongst the population[185]. What I do assert is that a more flexible notion of identity, not imposed by the state and recorded only temporarily and where necessary, would leave individuals more free to define their sexual status — if they chose to do so at all.

Conclusions

The categories “male” and “female” are pernicious not only in their unquestioned ubiquity but in their unnoticed effects. Kimmel writes of men and women that “by using separate facilities, we ‘become’ the gentlemen and ladies who are supposed to use those separate facilities. The phy- sical separation of men and women creates the justification for separating them — not the other way around[186]”. In other words, being told that we are different prompts us to act differently — and we then assume that we have been different from the start.

If we are to confront such “assumptions about the nature of difference”, Martha Minow tells us that we must question the point of view from which difference is assessed, recognise the subjectivity of difference as neither natural nor neutral, and acknowledge the perspectives of those assessed as different[187]. Such has also been the goal of this paper, and such has been the nature of its challenge. In the pages above, I have tried to show that although classification is central to thought, the way in which Québec’s classification of men and women constructs the reality of men and women is untenable. What is needed, then, is to rethink the way in which Québec engages — and other governments engage — in such classification ; a better, fairer approach would see the state cease to force men and women into ostensibly permanent and exclusive compartments, instead engaging in a nuanced and intervention-specific analysis of the varying perspectives yielded by the full continuum of sexual difference.

This proposal would not see “men” and “women” disappear as indices ; nor does it argue that men and women are the same. The human need to categorise and the persistence of sexual categories suggest otherwise — as do the numerous persons from all points along the sexual continuum for whom sexual identity is truly important. Rather, I assert that the human need to categorise need not be reified by the state. Just as “official classification [is a] political choice[188]”, Québec and other states can make a political choice to cease classifying its citizens permanently by sex.

What a difference that choice would make.