2. Mass Capture against Memory

In order to exclude Chinese migrants from the sphere of citizenship, the Canadian state laboured to remember these migrants through an elaborate system of capture that relied on an identification certificate known as a CI 9, which was demanded of each and every migrant who wished to temporarily leave Canada. The state sought to be able to recall each migrant who left the country in order to determine the validity of their right to return. In so doing, it created a vast archive for the purposes of exclusion. Yet, the mass capture of Chinese migrants to Canada in the first part of the twentieth century reveals the limits of conflating memory with storage.

As large and as complete as this archive is, it also offers a glimpse into the difficulty of accessing its contents. It is a difficulty that is both a problem of the present and the past. The discrepancy between the objectives of the CI 9 system of immigration control, and its effectiveness in achieving them, is a perfect demonstration of what Wendy Chun understands as the mutual exclusivity of memory and storage – or, as she puts it, how memory does not equal storage. Chun observes that computers have “erased the difference between memory and storage: users now store things in memory, rather than storing memory traces” (2016, 52). As she goes on to elaborate in a wonderful footnote, “Memory and storage are not the same things … Storage or stocks always look forward to the future. Memory is not static, but rather is an active process: a memory must be held – constantly recalled – in order to keep it from fading” (191). Although Chun refers to computational practices in the present, the conceptual distinction she makes between memory and storage is particularly pertinent for a system such as the CI 9. This system is premised on the ability to remember a particular migrant by storing their identity in a fixed location. Its use as a system fails precisely because it cannot accommodate the kind of dynamism that is central to the work of memory. This chapter investigates the ways in which mass capture works against memory in order to uncover the agential possibilities of persistent unknowability. Even though the state attempted to make each migrant knowable, its failure to do so illuminates how certain forms of forgetting are also a form of resistance.

Storage

Mass capture, in the form of CI 9s, is bound to questions of storage. How these certificates have been stored – both during the time of their circulation as active immigration files, and throughout their sequestration as records of a historical process of distinction and discrimination – tells us how they were used, how useful they were, and how they might be used now. On every certificate there are hole punch marks. Usually, there are two holes at the top. Sometimes, there are three or four along the sides.

These are, of course, the punch marks left by the filing system used to store the certificates. The CI 9s were kept in the most advanced document storage system of their time: the binder. First manufactured in Germany by Louis Leitz in 1871, the loose-leaf binder was perfected with the addition of a finger pull hole in 1911. This system of collecting slips of paper into a tidy whole was a “miraculous, order-creating contraption” wherein “the whole of the state emerges” through a “trinitarian formula of bureaucracy: ‘The state, the bureau, and the file’” (Vismann 2008, 133). Neatly stacked, one on top of the other, threaded through by a metal clamp, and bound together, every CI 9 was stored in this way for future reference.

This system of physical storage extends the idea of documentary mass capture as a grammar of action in the previous chapter. The mass capture of non-citizens through the documentation of their identity is characterized by a replicability of actions that is at the core of Philip Agre’s theory of surveillance as a form of capture. The form of the documents, which I will discuss shortly, as well as the way in which they were physically stored, signals how orderliness is intrinsic to the automation of this storage system. As Cornelia Vismann observes, the binder enforces its own order: “The old goal of avoiding the conflict that arises from scraps of paper lying around everywhere is replaced by an instrument, the self-binder. It puts an end to instructions on how to handle records. The file itself prescribes the necessary activities. Starting with the punch, its individual physical parts predetermine a clear order: punch, open, fix, insert, close” (2008, 137). Neatly bound by a clearly prescribed order of activity, the CI 9s were punched through and filed for a future where they could easily be retrieved and put to use.

This process of copying and filing is a form of control and owes a debt to systems of accounting developed in US and British colonial slave plantations. “The word ‘control’ itself comes from an accounting document: the countreroulle or counter-roll, a duplicate of a roll or other document, which was kept for purposes of cross-checking” (Rosenthal 2018, 2–3). The issuance of two CI 9s – one for the migrant, one for the state, so each certificate could be checked against the copy on file – is an example of immigration control grounded in the power of the file to police the movements of Chinese migrants. Indeed, as Caitlin Rosenthal points out, the idea of cross-checking first meant “‘verification,’ but by the late sixteenth century it had come to encompass the direction, management, and surveillance that verification required” (2018, 3). The CI 9s that were kept on file by the state to enable this system of cross-checking are the counter-roll of an elaborate visual and textual record of control rendered through the technological innovation of ringed binders and filing cabinets.

It is because of the miracle of this filing system that researchers like you and me, more than a century after the creation of these documents, can consult this archive confident in its coherence. The camera alone is not sufficient. It must, as John Tagg recognizes via Gilles Deleuze and Félix Guattari, be tethered to a larger apparatus of capture: “This is how the machinery of capture works. To the magical capture of the image is harnessed the mechanics of subjection of a bureaucratic system, comes joined to the storage and retrieval system of the filing cabinet: the One-Eyed Man and the One-Armed Man, the two modalities of power that join in the technical-mechanic enslavement of the modern State” (Tagg 2009, 3). This hybrid creature works to expand the reach of the camera and to make sense of its processes. Merely having an image is not enough. The state had to be able to corroborate that image with a range of other information. What is more, merely putting image and text together is nothing without also having a system of storage that would make retrieval efficient and accessible. For the CI 9s, the Canadian government settled on a system in which the certificate number, above all the other information collected on each CI 9, would guide the filing and storage of this archive.

As I have noted, each CI 9 has been kept in meticulous order, sequenced by the certificate number and, following from that, the date of issue. However, this impressive legacy of orderliness also betrays a weakness. The dependency of the system upon the certificate number at the expense of any other information on the certificate indicates a problem of ordering, and, consequently, a problem of memory.

There is no question that the certificate number rules supreme in the organization of these certificates. The certificates’ arrangement by number not only dictates the way they were physically filed in the moment of their creation, but also how they were preserved on microfilm reels in the middle of the twentieth century. This organization of the materials constitutes a problem of memory in that it would not have been possible to find a certificate by any attribute other than the certificate number and its physical place in the binder or on the microfilm reel. Despite the promise and premise of retrievability, this information is not particularly useful for finding anyone, even though the very existence of an identification document seems to suggest the ability to do just that. Think about someone you know well. Do you know their passport number or driver’s licence number? We tend to remember people by their names, where they came from, what they do for a living, or what they look like. This remembering is itself a form of intimacy. The state does not require such an intimate level of recollection and so we are issued numbers for our identity documents. But, as I have been arguing, that process, on a mass scale, begins with the desire to identify those who do not belong, those who should be excluded. The ordering of these migrants by certificate number reduces them to these numbers. However, this reduction of a person to a number does not, in the end, result in practical efficiency. It is a show of efficiency, yes, but as this chapter will demonstrate, reducing memory to an act of storage is to risk the failure of memory. Such a reduction of a person to a number, of memory to storage, leaves this system of state memory vulnerable to failure. Indeed, memory falls apart at the level of storage.

Part of the failure of storage as memory in the case of the CI 9s lies in the resistance of Chinese names to alphabetical order. In her consideration of the rise of the binder as the preeminent system of document storage, Vismann notes, “Two worlds coincide in the binder: the mechanized world of the ordering apparatus and the alphabetical world of letters” (2008, 132). In the case of the CI 9s, the certificates certainly submitted to mechanical ordering – to being punched, fixed, inserted, and filed – but they were resistant to alphabetical sequencing. Although each migrant’s name was recorded twice (and sometimes thrice in the case of aliases), in English and in Chinese, this information was not part of the system for filing and ordering these documents.

Each name, twice or thrice recorded, escapes the world of the binder as storage. Names are irrelevant to the ordering of these certificates and the way in which they have been filed and stored. Even though names are putatively central for identifying a person, they are not the primary mode by which the Chinese migrants in the CI 9 archive have been remembered. Knowing a name is, and would have been, insufficient for locating the identification certificate connected to that name.

Although this insistence upon identification by number is dehumanizing in effect, it is also indicative of the way in which these names, and thus these identities, overwhelmed the system designed to capture them. The filing of identification photographs by number instead of name was not unique to immigration processes. Mugshots, organized into various “rogues’ galleries” as police exhibits or albums, were also filed by number. Thomas Byrnes’s famous 1895 book of criminal mugshots, Professional Criminals of America, featured mugshots set into grids of nine photographs per page (in rows of three) and were organized by numbers assigned to each criminal. It was easier to collect mugshots than it was to organize them in such a way that the identification information they provide could be easily retrieved. “The deployment of photography to document the criminal therefore posed a central problem: as the numbers of records increased, the utility of the collection decreased” (Finn 2009, 10). There were 204 mugshots in Byrnes’s collection, and by the end of the nineteenth century, “rogues’ galleries had grown in size from the hundreds … to the thousands, leather-bound books had given way to filing cabinets, and the resultant archives required the assistance of three attendants” (ibid.). Even though these mugshots may have been filed by number, like the CI 9s, the names of each criminal continued to matter in the way in which these files circulated.

Reference to criminals tended to privilege their names over their criminal identification numbers. As Lourdes Delgado argues in her study of the criminal mugshot as a site of a semantic history of guilt, even with the introduction of identification numbers for detainees, the number is listed almost as an afterthought or a way of lending administrative heft to the organization of police records: “From 1878, when referring to a criminal, the New York Times began to include a number, referring to the police archive, where the records of detainees were held, with the subsequent suggestion of scientific, or at least administrative, certainty that the numbers provided” (2017, 223). Thus, for example, the Times would report on “James Williams, alias John Lestrange, a pickpocket, (no. 1, 132 in the Rogues’ Gallery)” (quoted in ibid.). This emphasis upon the name of each criminal as part of their identity was part of a larger effort to make the individual criminal familiar and knowable. One of the reasons why Byrnes assembled the albums for public circulation was to address the challenge of making the photographic records useful by “enlist[ing] the public in the practice of policing” (Finn 2009, 10). Although the mugshots are organized by number, their circulation is meant to make them more, not less, indexed to a person whose identity could be differentiated and distinguished by those who would put in the effort to look at them closely.

In contrast, the CI 9 system favoured differentiation and distinction through an identification number first and foremost. Upon returning to Canada and requesting re-entry, the migrant would present their certificate and it would have to match the one on file before they were readmitted. The filed certificate was retrieved via its place in the numerical sequence. Filing and storing each certificate by number rather than name displaces the burden of proof of identity onto the documents themselves. In relying so heavily on a specific document and document number, at the expense of all the other identification information, the state invests in the reading of documents rather than the reading of bodies.

The information about these bodies is certainly collected, but it is not used as the principal method for differentiating one migrant from another. All of the other information, recorded and kept with such diligence, merely corroborates what the certificate already tells: that this person is who their documents claim them to be. The representation of the racialized subject in an identification photograph calls attention to the dangers of misrecognition and the instability of the connection between the photograph and its subject. For the viewer, any identification photograph poses the silent question: Is that you? To request an identification photograph is to question someone’s identity.

These are images that the law both demands and rejects. It demands the photograph as proof of the subject’s identity. But it rejects it as questionable evidence at best. For there is a curious logic to demanding photographs to identify racialized migrants when another, contrary logic declares that they all look the same. Anna Pegler-Gordon notes that Congressional representative Thomas Geary, the primary sponsor of the United States’ Chinese Exclusion Act (1892) and its amendment in 1893, argued in support of supplementing repetitive textual descriptions with photographic documentation of Chinese immigrants because “all Chinamen look alike, all dress alike, all have the same kind of eyes, all are beardless, all wear their hair in the same manner” (quoted in Pegler-Gordon 2009, 35). Like Sir James Herschel in India, whose reliance on fingerprinting I will discuss in chapter 4, Geary found textual descriptions to be inadequate for distinguishing between racialized subjects, but unlike Herschel, Geary turned to photographic identification as a way of resolving this problem. However, Geary’s belief in the power of the identification photograph to distinguish between racialized subjects highlights the contradictory logic behind the demand for identification photographs from supposedly indistinguishable racialized subjects.

In this way, the document, the certificate, comes to stand in for the non-citizen. The body is secondary. It is further proof, a form of corroboration, but it is not the first site of examination. Files not only come to stand in for the body at the site of inspection, but also begin resembling bodies. Vismann herself begins to think of files as being like people: “Once files act like people by staying overnight and spending extended time in corridors, that is, once files take on the habits of their users, a new time has begun – the time of files” (2008, 139). Vismann’s whimsical anthropomorphization of paper files is particularly poignant given the role and use of files during the Weimar Republic. To have a file, to be filed, also had horrific consequences for those who were persecuted by the Third Reich. Think of the chill experienced by those who realize that state intelligence services might have a file or dossier devoted to them. For Vismann, the time of the files is one of increased automation in terms of the management of information. But Vismann’s provocative formulation can be easily extended to fall within the ambit of mass capture.

To imagine the files as people who spend the night in the binders and cabinets of their storage is to begin to see how these seemingly innocuous office supplies offer a way into understanding the files as an extension of the hold of the ship and the barracks of the migration detention centre. Bodies must be processed and housed just as the information about those bodies must be acquired, transcribed, and then punched through, inserted, and filed. Vismann’s contemplation of files that act like people calls to mind Allan Sekula’s engagement with the body and the archive, or the way in which “the camera is integrated into a larger ensemble: a bureaucratic-clerical-statistical system of ‘intelligence.’ This system can be described as a sophisticated archive. The central artifact of this system is not the camera but the filing cabinet” (1986, 16). As Sekula understands, given the sheer mass of information that the photograph provides, the photograph must be “tamed” through an apparatus, the filing cabinet, that will allow it to be retrieved (17–18). In the case of identification documents such as the CI 9s, the state does not just want the files to act like people, it wants them to be like people. For the state to make non-citizens, the capture and containment of these files, files threaded through the metal clamp of the binder, is coextensive with corporeal mass capture.

Yet these bodies and these files remain stubbornly resistant to recall. Stacked one upon the other in containers that dictate the precise dimensions of the space that they are permitted to occupy, the files that come to stand in for the bodies that they represent offer a promise of order and containment. However, files are not bodies. They may sometimes act like them; they may even, as in the case of the CI 9s, stand in for them. But the representation of the person is very clearly never the person as a whole. Nonetheless, the CI 9 system continually mistakes the file for the person. In so doing, it misunderstands storage as memory.

This investment in files and numbers over language not only fetishizes the document over the person it represents, it also attempts to reduce the act of reading a body to that of matching one number with one another. Through faith in numbers, the turn is not just away from language itself, but specifically away from Chinese names and the Chinese language. It might have been disdain for Chinese and its perceived archaism. As Thomas Mullaney observes in his study of the invention of the Chinese typewriter, among Western thinkers (most notably Hegel), ideographic languages such as Chinese, Incan, and Egyptian were “understood as being stuck in a state of arrested development that, in turn, froze in time those who spoke and thought in, with and through these languages” (2018, 66). Even well into the late twentieth century, “concepts of alphabetic supremacy and Chinese linguistic unfitness were not so easily dispelled” (69) such that “questions of Chinese linguistic unfitness … moved out of the politically untenable realm of race, and into the sanitized realm of technological devices like the typewriter” (70–1). Against the technological backwardness of the ideographic script, the CI 9s conflate an identification number (the certificate number) with the identity of the person. It goes without saying that a person is far more than the number to which they have been indifferently assigned. But, in refusing to engage in the complexity of Chineseness by reducing Chinese identities to thousands of five-digit numbers, the state fails to remember the very people it seeks to track.

Failed Memory

Memory is not storage. To think of one as interchangeable with the other is to risk the failure of memory. For mass capture, failed memory means the failure to accurately identify people who are represented on the CI 9 certificates. This system is premised entirely on the ability to remember a particular migrant by storing their identity in a fixed location. Its use as a system fails precisely because it cannot accommodate the kind of dynamism that is central to the work of memory. The CI 9s unsettle memory not least because even the most basic attempt to use these certificates for identification illuminates the ways in which they completely foil the identification process. The CI 9 system, for all its complexity and insistence on detail, could not handle the peculiarities of Chinese names.

Each migrant’s Chinese name was anglicized by the immigration agent, with the help of a translator, at the time the certificate was issued. This process of anglicization was, at best, improvisational. At the time, there were no conventions for anglicizing Chinese names. A surname as common as “黃” (Wong) could be written in English as Huang, Wang, Wong, or Weng. Chinese migrants spoke a range of dialects including Cantonese, Sze Yip, Toisanese, and Fujianese. To further complicate things, another very common surname, “王” (Wong), is a homonym for “黃” in Cantonese. Each dialect could have lent a different inflection to the anglicization of that particular name. Further, there was no distinction between surnames and proper names in the certificates. Following Chinese naming conventions, some agents entered surnames first; some did not.

All of these differences lead not only to the problem of accessing this history in the present, but also in the past. As the confusion over something as straightforward as names suggests, the CI 9s are also a testament to the ways in which massive projects of identification are so vulnerable to failure. See, for example, the certificate that inaugurates this new era of identification, certificate number 15811 (fig. P.1). Here, so this document claims, is Luie Lee Teck. Maybe? The handwriting is hard to decode. He has written his name in Chinese closer to the bottom of the certificate. But somehow this additional information only makes things more confusing. The Chinese writing is clearer, but the sound of the characters does not correspond in any way with the English transliteration on the form. In Cantonese, these characters would sound something like dak shi lum. Even if the man in this certificate spoke a dialect other than Cantonese, the opening consonants of each character of his name would not be that different. But then something twigs. That first part of the transliterated name might be “Lum.” Perhaps, then, the name has been phonetically transcribed so that the surname has been replaced by a character in the given name? Still, there is uncertainty.

We do not even know his name. Or rather, we have two names and we cannot know if one, or both, or neither, might be correct. The textual material is in contradiction with, or belies, the claims of the visual evidence or photograph. In The Disciplinary Frame, John Tagg highlights the ways in which repressive state photography, such as identification photography, was pointless without a coherent system of storage: “It was no use accumulating records if the storage system did not make it possible to retrieve them, cross-reference them, or compare them” (2009, 13). Tagg’s point converges with the skepticism of scholars of memory studies regarding the idea of memory as storage. Notably, Paul Connerton (2008, 65) identifies a surfeit of information as a key form of forgetting. Even though the state attempted to remember individual Chinese migrants by creating an apparatus of information capture in the form of the CI 9, the actual result was less an active process of remembering than an imperfect system of information storage. It was imperfect not least because the system of retrieval could not keep up with the sheer scale of the information that had been captured and stored. As Nancy Van House and Elizabeth Churchill (2008) emphasize, the design of systems of storage and retrieval fundamentally shapes the memories that can be retrieved. For example, there was no retrieval system for names in Chinese. The certificates made a point of capturing both the anglicized and Chinese name of each migrant, but this information was useless for the purposes of retrieval because the certificates were stored by number.

Not only was it difficult to recover information from these certificates (without a certificate number), but my research team found that significant mismatches between the English and Chinese names on the certificates reveals a strong possibility that the names themselves were practically useless for identifying migrants. In an analysis of 112 certificates issued in 1913, my team found considerable differences between the Chinese names that the migrants provided, and the names as they were anglicized by the immigration agent.

Fig. 2.1 Name correlations in CI 9 certificates numbered 23500 to 23355

Figure 2.1 provides a summary of our findings. As you will see, there were very few cases (less than 1 per cent) where the Chinese name and the anglicized name corresponded. Sometimes there were slight differences (what we term “low difference”). A third of the time, we found moderate differences. And, in a finding that I still find surprising even though I suspected that there would be many such cases, almost 60 per cent of the Chinese names significantly differed from the anglicized ones on each certificate. Each certificate was examined and transcribed by at least two research assistants working under my direction. Wherever there were major differences in transcription, we consulted each other and worked together to arrive at the most likely name. We took into account the fact that the certificates did not differentiate between surnames and given names, that Chinese surnames are often written and given first, and that the Chinese names were often written according to the linear protocols of Chinese script so that the written Chinese names must be read from right to left. We also converted the names to Cantonese pinyin and tried variations of other regional dialects including Toisanese. Even after taking all of this into account, we found an enormous number of discrepancies between Chinese names and anglicized names on the certificates we examined.

For this chapter, we chose to work with a sample of certificates from 1913 because the iteration of the CI 9 system requiring photographs began in 1910; we wanted to examine certificates that were produced a few years after the inauguration of this system in order to see how it functioned after it had a chance to stabilize. The fact of the differences, the scale of them, is telling. If nothing else, these certificates suggest that the identities of the men, women, and children on these certificates could not be taken for granted. These identities certainly circulated differently depending upon the linguistic register.

That Chinese migrants sometimes used false documents in order to gain entry into Canada, particularly during the period of exclusion (1923–47), will come as no surprise to any observer of immigration history. Chinese migrants avoided exclusion by sharing head tax papers in order to be “readmitted” to Canada when in fact they were entering for the first time. Contemporary Chinese-Canadian fiction such as Wayson Choy’s The Jade Peony, Sky Lee’s Disappearing Moon Cafe, and Judy Fong Bates’s Midnight at the Dragon Cafe, as well as nonfiction work such as Choy’s Paper Shadows, have made the idea of the “paper son” part of the parlance of Chinese immigration. My research supports the stories told by these writers and indicates that paper sons, and from them entire paper families, are not exceptional. These are not singular or occasional incidents. It was not just one or two people declaring themselves to be someone they very likely were not.

That we still have tens of thousands of perfectly ordered CI 9s, all filed neatly by number and date of issue, each bursting with a wealth of information, is a credit to the extensive reach of this system. However, as I have been arguing, this system of collection was only as effective as the system of retrieval – and here, the system fails. The expectation was that each migrant would return to the intended port of re-entry where their certificates would be kept on file for cross-referencing. That, of course, was the first obstacle. Entry and re-entry, while ostensibly dependent on an easily filed and retrieved CI 9, also depended on the immigration agents. Fraud could be enabled by the shortcomings of port guards who were invariably patronage appointments, including, as the 1910 Royal Commission to Investigate Alleged Chinese Frauds and Opium Smuggling on the Pacific Coast details:

W.A Kent, aged 48, given to drink and unreliable.

T. Physick, aged 63, physically unfit and service unsatisfactory.

H.H. Warburton, 50, cripple and unfit.

J. McPherson, given to drinking and unreliable. (Mar 2010, 43)

As Lisa Mar observes, “Even the immigration officials’ measuring stick for height was discovered to be inaccurate, with one foot being only eleven inches” (114). Thus, this system of identification, information retrieval, and cross-referencing depends upon a series of agents who were themselves susceptible to a wide range of human errors. Despite these weaknesses, there was much hope placed in the potential of photographic technology. Mar notes that the “absence of photographic documentation [prior to 1910] encouraged fraud” (ibid.). And yet, as I have been suggesting, even the turn to photographic identification could not prevent fraud. If anything, the systematic investment in the indexicality of the identification photograph, the idea that the image presented would correspond seamlessly with the text on the page which could then be unproblematically indexed to the bearer of the certificate, makes this system even more vulnerable. In this system of mass capture, the photographic turn exacerbates the failures of relying on the indexical. Mass capture manufactures non-citizens by making them hyper-visible to the state, but this process is also deeply vulnerable to failure.

Bodies beyond the Archive

What emerges out of the failures of mass capture? In the first part of this chapter, I have shown that mass capture as a technology of non-citizenship fails. It fails because so many of its subjects remained elusive and largely invisible. This failure lies not only in the unwieldiness of the system, nor even in the incompetence of those charged with executing its grammars of action. Of course, these are factors. But the failure of the CI 9 system poses its greatest threat to the ideological underpinnings of surveillance: the CI 9s persistently undermine the idea that a photograph of a person can necessarily be counted upon to be a representation of that person.

As a collection and a specific archive, taken together, the CI 9 photographs tell us that identification photographs cannot simply be read as such; that repressive photography, as tracked by Sekula and Tagg, might also serve as a site of escape. Recall that Tagg sees in identification photographs the “traces of power, repeated countless times, whenever the photographer prepared an exposure” and that the “repetitive pattern” is central to the repression of the subjects of identification photography: “the body isolated; the narrow space; the subjection to an unreturnable gaze; the scrutiny of gestures, faces and features; the clarity of illumination and the sharpness of focus; the names and number boards” (1988, 85). Still, as Sekula notes, proponents of identification photography were well aware of its limitations. Their response to these limitations, as shown by Sekula’s now canonical overview of the methods of Francis Galton and Alphonse Bertillon, were to imbricate the photographic representation with the statistical. What these responses fail to account for is the possibility that there might not be any correspondence between the photographic subject and the biographical, biometrical, and statistical information collected to correspond with that subject.

The body in the archive is not necessarily the body that the archive believes it to be. Certainly, there is a body. But whose? In the CI 9 archive, the bodies belonging to these names have moved far out of the disciplinary frame. What is more, they have used the frame itself as a cover for their escape. Take, for example, Wong Lin Foon, CI 9 number 22711 (fig. 2.2). According to this CI 9, he registered to leave Canada on the SS Cyclops on 1 October 1913. Given Tagg’s observation of the machinery of capture as being akin to man with a single eye and arm, how fitting it is that he should have shipped out on a ship named after a mythical, one-eyed creature. In so doing, he gave the apparatus of mass capture the slip. Filed just in front of his certificate is a letter (fig. 2.3).

Exploiting the vulnerabilities of the system, Wong Lin Foon claimed to have lost his identity documents. Those documents were located and found to have belonged to someone else entirely, but by the time this ruse was uncovered, he had sailed on the SS Cyclops. He successfully registered out using CI 9 no. 22711, and thereby not only engineered his own exit with false documents but also seemingly allowed at least one other migrant to leave with documents that were not his own.

In a further twist, there is another CI 9 no. 22711 (fig. 2.4). This one appears to belong to a man named Ing Yon Lan. Let us read the story of these three documents: the two CI 9s claiming to be no. 22711, and the letter that is filed with the one belonging to Wong Lin Foon. First, the chronology.

On 1 October 1913, Wong Lin Foon departs on the Cyclops from Victoria, British Columbia, Canada, by registering his outward-bound migration on CI 9 no. 22711. On that same day, he claims to have lost his CI 5 no. 39222 and applied for a CI 28, which is an affidavit detailing his identifying information used to replace the lost CI 5.

On 8 October 1913, Ing Yon Lan departs Vancouver, British Columbia, on the Empress of Russia. In lieu of a CI 5, he has filed a CI 36 which was issued to replace CI 5s that had been issued prior to 1912 when there were no photographs attached to these certificates.

On 11 September 1914, the chief controller of Chinese immigration, William Duncan Scott, responds to a query from an immigration agent in Victoria dispatched on “the 4th instant,” or 4 September 1914. In this letter, Scott uncovers a lie. Wong Lin Foon is not who he claims to be. Or, he has not lost the head tax certificate he was issued upon entering Canada, CI 5 no. 39222. That document, Scott reveals, “evidently belongs to a party in Vancouver.” Although Scott does not disclose how he came to know this information, one gathers that the owner of CI 5 no. 39222 did not board the Cyclops. At the time of his missive, according to Scott, Wong Lin Foon has not yet returned from China. Scott assures the agent in Victoria that an investigation will follow upon this man’s return and that the agent should wire Scott should he wish Wong Lin Foon to be detained.

On 24 September 1914, less than two weeks after Scott’s letter was dispatched, Wong Lin Foon returns to Canada. Seemingly, he has done so with little fanfare. The agent at the port attests, as the form dictates, that he has “personally examined the person of Chinese origin who claims to be the person above described and whose photography is affixed hereon … and declare him to be the same person” (emphasis mine). Wong Lin Foon, or someone claiming to be Wong Lin Foon, countersigns the CI 9 on the lower right-hand corner. And that is that.

On 30 September 1914, Ing Yon Lan returns to Canada by landing at Victoria. His return date is noted and stamped, and scrawled across the top right corner of this CI 9 are written the words: “Released at Victoria 30/9/14.”

It does not seem that either man was detained upon entry. The best indication of the unproblematic nature of their re-entry into Canada can be found on the CI 9s themselves: both certificates look, more or less, like every other one in the archive. True, Wong Lin Foon’s was counter-signed by both the re-entering migrant and the controller at the port. In contrast, Ing Yon Lan’s was not. Still, nothing indicates a problem with the identification photograph itself. In the case of Wong Lin Foon’s, it has been neatly initialled by the controller. It does not seem to have been tampered with, and moreover, the photograph had been on file all this time, so the original could not have been swapped out with another. Neither certificate was cancelled, crossed out, ripped up, or treated with any kind of exceptionality.

And yet, Scott’s letter was eventually filed with this certificate so that I would find it there, tucked neatly away, alongside the certificate question. At some point, someone knew that these two documents belonged together. But not on 24 September 1914, when Wong Lin Foon returned.

Once again, it is unclear that the officials have captured the names of the migrants correctly. In the case of Ing Yon Lan, the Mandarin pinyin of the characters he writes on his form, 連有伍, would be “Lian You Wu,” and the Cantonese pinyin would be “Lin Jau Ng.” In the case of the Cantonese pinyin, the differences are not so enormous between the name as it might have been pronounced and as it is transcribed. But even minor differences reveal the challenges faced by this system. In the case of Wong Lin Foon, his signature upon departure places the character for Wong (王) last even though it is written in English first. To complicate things further, when he returns, he writes his name vertically and this time Wong comes first. Given the overwhelming popularity of Wong as a surname, it seems likely that Wong is the surname intended for this document. That is: it seems likely, but we don’t really know.

So, someone claiming to be Wong Lin Foon departed Canada. And someone claiming to be him, or someone who looks enough like him to resemble the identification photograph on CI 9 no. 22711, returned to Canada a little less than a year later. In light of Scott’s letter, the man claiming to be Wong Lin Foon was probably not Wong Lin Foon. There are now already two bodies attached to one identification certificate (CI 5 no. 39222). The man claiming on his return to be Wong Lin Foon on CI 9 no. 22711 might or might not have been the same person who departed Canada using that certificate. Thus, there might be a third additional body attached to this certificate.

And then, just as importantly, there is Ing Yon Lan, who is, according to his photograph and the biographical information provided, an entirely different man. A man who may or may not have been him certainly left Canada in 1913, and a man who may or may not have been the man who left entered Canada in 1914 on the basis of that certificate. Have you been counting? I am beginning to lose track. The evidence found in the two CI 9 nos. 22711 indicate that the possibility that as many as four different men could have travelled to and from Canada using a single certificate number.

CI 9 no. 22711 gives us as the stories of at least two men. Wong claims to be thirty-six years old and a labourer in British Columbia. Ing claims to be a launderer who is twenty-six years old and living in Montreal. Both happen to be the same height, 5’6 ½”, but their photographs reveal their facial differences. Somewhere, in between one certificate and another, between one set of a details and another, there might be the truth of one or another identity.

There is one more element of CI 9 no. 22711 to take up. It is not the only double. There are actually thousands of doubles. In a move that remains astonishing to me, this duplication is deliberately built into the system. Note that Wong Ling Foon departed from the Port of Victoria. Now look and note that Ing Yon Lan left via the Port of Vancouver. The CI 9 system used one set of numbers for certificates of migrants departing from Vancouver, and then used the same numbers all over again for those leaving from Victoria. Here is a breakdown of the CI 9 numbers for each port of departure and the LAC reel numbers in which each certificate can be found:

Vancouver departures

T-6047 – Sept 1910 to Jan 1912 – CI 9 no. 17286 to 20110

T-6048 – Jan 1911 to Oct 1913 – CI 9 no. 20111 to 22899

T-6049 – Oct 1913 to July 1916 – CI 9 no. 22900 to 27772

T-6051 – Nov 1917 to July 1919 – CI 9 no. 28650 to 31399

T-6052 – June 1915 to Sept 1929 – CI 9 no. 1 to 03240

Figures 2.13 to 2.20 are the Vancouver doubles of the Bellerophon certificates I discuss above. Each one of these certificates shares a number with another. They are differentiated by the ports of departure, either stamped or written into the first blank space on the form. Given the primacy of the certificate number on these certificates, it is surprising to say the least that the entire archive holds thousands of these doubles – certificates that share one number but were issued in different cities to different migrants. Threaded throughout the archive, these doubles were not considered a deficiency in the system. The Department of Trade and Commerce seems very confident that having one set of certificates issued from Victoria, and another set issued from Vancouver using the very same numbers, would not pose any kind of problem for the management of these files. Still, even if the certificates from Vancouver and Victoria were always perfectly siloed, even if they never crossed or mixed, these doubles suggest that the primacy of the certificate number as a central part of document retrieval is flawed. The number alone does not suffice. It never did.

In addition to the problems attendant upon relying on certificate numbers, there is also the issue of the names themselves. At least in the case of Wong Lin Foon, the characters correspond to the transliteration. As I noted earlier, there is often no correspondence between the name in English and the Chinese name. For example, there is CI 9 no. 23413 (fig. 2.21). His name is given in English as “Yee Chan Kam.” He writes his name in Chinese as “奎贊余.” The Mandarin pinyin for these characters is “Kui Zan Yu.” The Cantonese pinyin (and it is much more likely that he spoke Cantonese or a dialect more closely connected to Cantonese) is “Fuk Chan Yu.” What are we to do with this riot of difference? The lack of correspondence between the anglicized name and the one in Chinese script is an indication, as I have been noting throughout this chapter, of one of the many ways in which the CI 9 system fails to capture Chinese migrants. But there is more here than just the fumblings of a state system that could not get a lot of names right. There is the possibility of agency on the part of non-citizens.

I do not wish to overstate the forms of agency available to the subjects of the CI 9s. As Michelle Caswell cautions, in the context of reading mug shots produced at Tuol Sleng or S-21 prison in Cambodia, “we must tread carefully for fear of conferring a false sense of agency on the part of the victims” (2014, 59). I recognize that the conditions under which the CI 9s were produced are hugely different than the conditions at Tuol Sleng, where “the subjects of these mug shots did not choose to be photographed, they did not have any ownership over the way they were depicted, they were acted upon against their will” (59). The CI 9 system is a far cry from the “total institution” Caswell describes (52). Still, Caswell’s caution against a “concept of co-creatorship [as] a fallacy that grants too much agency to the subjects of these photos” remains important when attempting to find agency in oppressive photography (59). The Chinese migrants in the CI 9 photographs were not agents or co-creators of the CI 9 archive. However, as the discrepancy between Chinese and anglicized names indicates, there is the possibility of small acts of refusal.

Understanding that a significant portion of the head tax certificates in this archive are, at some level, fictional, given the scanty correspondence between the Chinese name and its anglicized counterpart, the CI 9 certificates are not so much an archive of state repression as of subaltern ingenuity. I am not necessarily suggesting that migrants deliberately sought to delude the controllers. Certainly, some might have. However, I am interested in the possibilities of more passive forms of dissimulation. A lapse here; a decision not to correct a mistake there. A shrug of the shoulders when one realizes that the bureaucrat on the other side of the table has bungled one’s name beyond all recognition. These are small resistances that nonetheless accrue. They accumulate so that, over time, an entire system, and the archive that accompanies it, can no longer be trusted.

In the cases of Wong Lin Foon and Yee Chan Kam, at least as they are known in the CI 9 archive, the photographs do not index the men that the certificates claim to represent. They, along with many others, undo the bind between identification photograph and the person identified. These photographs put into question not only the claims of the individual certificates themselves, but also those of the entire project of photographic identification. By injecting so much uncertainty into this system of mass capture, the subjects of the CI 9 photographs insistently undermine the work of capture.

Again and again, certificate after certificate, Chinese people in this archive refused the demand to be made known and knowable for the purposes of being excluded from the sphere of citizenship. The CI 9s are evidence of unknowability as one form of agency in the face of the deprivations of non-citizenship. If mass capture as a technology of non-citizenship is premised on making the non-citizen knowable to the state, this study uncovers instead a project of mass misrecognition and invisibility. In the CI 9 archive, thousands of Chinese migrants hide in plain sight.

It is not clear whether or not the state had a sense of the scope of the failure of this project of capture. It is not only the state that invested some belief in the truth-claims of head tax certificates. For example, as suggested by projects such as the Chinese Head Tax Searchable Database, developed by a team at the University of British Columbia, head tax information has been seen as one way of conducting ancestry research. That database is based on the Chinese Head Tax Register, which is similar to a ledger book that contains information on Chinese migrants transcribed from various certificates, but it does not contain the certificates themselves. Therefore, there are no identification photographs, and, given that the information in the register is copied from individual certificates, there is an even greater chance of errors than on the CI 9s.

Given the results of my research, I would caution against relying too heavily upon materials such as the head tax register as a site for genealogical research. Certainly, the CI 9s are an incredibly important and rich source of genealogical information. I do not want to take away from the importance of this information. What is more, for Chinese-Canadian people who are searching for family, given the histories of racism and persecution that these families survived, the CI 9s offer more information about individual Chinese migrants in Canada than any other set of documents. People have found their families through the CI 9s. Still, although there are many names, and many identities, it is not always the case that these are real names attached to real people. Rather, these identities should be read along a broad continuum with truth on one end and necessary fictions on the other. That there was fraud is not in doubt, and the kinds of fictions that are the basis of these frauds should be a celebrated part of the story of the CI 9s. For Mass Capture, these fictions are instances of the kind of ingenuity and creativity exercised by people who often had very few resources for escaping the strictures of this discriminatory system of capture and exclusion.

Of course, the state had some sense that its attempts to prevent fraud were not successful. As Laura Madokoro observes, “From the late 19th century, fake identity documents and elaborate coaching papers enabled Chinese migrants to adopt new personas and gain admission under false pretexts. Officials in white settler societies were well aware of the practice, but in the face of language barriers and brokers who benefited financially from the schemes (such as translators whose services could be purchased to facilitate entry), authorities found themselves at a loss as to how to effectively combat these practices” (2016, 98). While the Canadian government was stymied in its attempts to prevent fraud, it did attempt to punish the fraudulent. In the late 1950s, the Royal Canadian Mounted Police (RCMP) began investigating immigration fraud committed by Chinese people in Canada. “Upon confirmation of fraud in Hong Kong, the RCMP, aided by plain clothes officers from the Royal Hong Kong Police Force, raided Chinese homes and businesses across Canada. Their targets included the Chinese Benevolent Association in Vancouver, which the RCMP alleged was overseeing the paper son scheme” (119). However, this attempt at punishment was limited. Chinese community leaders lobbied intensely against what they saw as forms of discrimination based solely on race. Further, “the costs of enforcement were politically and economically high enough and the phenomenon of illegal entry sufficiently widespread to convince Canadian authorities that prosecution and deportation were not viable solutions to the problem of illegal entry. Instead of punishing paper families, the Canadian government introduced a Status Adjustment Program in 1960 whereby all those who had moved to the country illegally could confess and then apply to be naturalized, receiving all the rights of other naturalized Canadians, including family sponsorship” (ibid.). This program was, as Madokoro points out, consciously modelled after the Chinese Confession Program in the United States.

After decades of legislated discrimination and exclusion, thousands of Chinese people living in North America overcame their fears of exposure and deportation and identified themselves as illegal migrants. They did so through a formal process of confession. However, amnesty was not a guaranteed outcome, given that “the state-proffered solution” left Chinese migrants “susceptible to the subjective operation of power: they were accused and interrogated and their possible futures in the United States and Canada ultimately depended on the discretion of immigration officials” (Madokoro 2016, 121). Madokoro reports that 13,865 people in the US took part in the confession program. There was also a significant number of participants in the Canadian program. “By late 1966,” notes Patricia Roy, “10,533 Chinese had applied for adjustment, 9,962 had been processed, and the department expected to clear the backlog in a few months” (2007, 298). But there is a difference between applying, being processed, and being made a citizen. It is not clear how many of those who had been “processed” were granted citizenship.

Even though there appeared to be widespread participation in these programs, we do not know “how many hundreds or thousands of other migrants never came forward to have their legal status adjusted, or more accurately, readjusted by the state” (Madokoro 2016, 121). The necessarily voluntary nature of the confession programs, and the inevitable failure of such programs to capture every non-citizen, signals ways in which many bodies in this archive remain inscrutable. The confession programs undoubtedly went a long way towards remediating and normalizing a number of fugitive subjects and fugitive families. However, we will never know the full extent of the failure of the head tax system, and in particular the CI 9 system.

How, then, to understand the failure of mass capture in terms of the making of non-citizenship subjects? I have been suggesting that mass capture is a technology that produces non-citizens. I have identified the CI 9 certificates as a prime instance of mass capture. Now, we are faced with a broken or corrupt technology. The archive is not a machine of repression, but rather, to recall the magical one-eyed and one-armed creature invoked by Tagg, a beast that is half-blind and maimed. These thousands of bodies that have moved well beyond the archive indicate not simply a failure of mass capture. Rather, and more importantly, they show how mass capture is productive of forms of agency and subjectivity. Non-citizens are not wholly figures of abjection and loss. They are not necessarily the subjects of repression and state capture. These thousands of mismatched names and bodies suggest that the non-citizen can be a figure of refusal and resistance. Even without the tenuousness of the connection between the textual information and the photographic material on each certificate, the CI 9s offer a powerful testament to the refusal of the demand to be documented.

Constructing Inclusion out of Exclusion

Although the CI 9s were a technology put into place for the purposes of excluding Chinese migrants from citizenship, in the many instances of evasion and refusal that emerge out of this archive, I also see a demand for inclusion. In the final part of this chapter, I will outline the workings of refusal and resistance in terms that understand the photographs attached to these certificates as part of a different cultural formation: a Chinese view of photography, and the traditions of Chinese portraiture. As I have noted throughout this chapter, identification photography is never simply functional. Indeed, as Karen Strassler suggests in her discussion of identification photography’s use during the New Order in Indonesia, “the identity photograph became a widespread visual idiom for legitimate belonging within the state-authorized national community, but the state’s fetishization of ‘proof’ of identity also gave rise to doubt and irreverence about documentary truths. Moreover, when appropriated for popular use – circulated among friends, incorporated into personal albums, or reframed as memorial images – identity photographs enter into spheres of social relations and identifications that challenge the state’s claim to be the sole agent of recognition within the nation” (2010, 21). It is not only the state that has the power to bestow identity via the photograph. These photographs carry a power that exceeds that of their supposed utility. They were used by their subjects outside the frame of the CI 9 as sites of self-representation, and as a form which could be suffused with an entirely different set of cultural norms and visual cues.

 

[2] Outside the Mass Capture Frame

The CI 9 photographs were not simply captured and filed. They travelled. Chinese migrants brought their own photographs to be affixed to the CI 9s. At least one other copy of the photograph, the one held by the migrant who departed, would have been outside the confines of the binder and the file. But it is important to also imagine how they might circulate detached from the certificates. Freed from the textual context of the immigration certificate, these identification photographs offer the possibility of other uses and forms of circulation. Taken out of the frame of state documentation, many of these head tax images could certainly function within the tradition of honorific or bourgeois portraiture. As Christopher Pinney notes in his introduction to Photography’s Other Histories, “A greater sense of the fragility and instability of the relationship between images and their contexts might allow the exploration of why certain images prove capable of recoding while others are more resistant, and many others are completely intractable” (2003, 4). Unfixed from the strictures of the head tax certificates, these photographs could very easily have been disseminated among families and friends.

The size of the photographs also indicates the possibility that they were used for purposes other than identification. The photographs measure approximately 2½ inches to 2⅜ inches by 3 inches. The CI 9 forms allowed for a space of 2½ inches by 4 inches. These measurements suggest that many of these photographs could have been taken with standard carte de visite cameras, which allowed for up to eight exposures. The cameras that produced such images were inexpensive, popular, and thus readily available. The carte de visite cameras also promoted the construction of images of their subjects as prosperous and confident through the use of props and costumes. As Mary Warner Marien notes, these cameras “encouraged sitters to construct an image of self-satisfaction and financial prosperity” through the use of “fancy furniture and painted backdrops” (2002, 84). The possibility that these identification photographs could have been taken inexpensively while offering elaborately constructed forms of self-representation, and multiple exposures, strongly suggests that many of the photographs affixed to the head tax certificates would also have circulated as portraits used for private consumption. Indeed, given that the state would have required only one or two images, it seems entirely likely that there were surplus images in circulation.

It would be a mistake to think of these photographs merely as responses to a governmental demand for identification. Identification photographs are rarely simply functional. They move across spaces and genres. Jennifer Bajorek, with reference to national photography in Senegal, writes of “the impossibility of deciding on and maintaining distinctions between genres” (2010, 159). A Senegalese photographer tells Bajorek, “‘We all did ID cards, and we all did portraits’; ‘All of us who did portraits also did ID cards’” (ibid). The distinction between the portrait and the identification photograph is both maintained and collapsed in this photographer’s declaration. He did both. In doing both, he suggests that the contrast between one and the other is a matter of degree rather than absolute difference. As the work of Bajorek, Finn, and Strassler shows, the identification photographs, once they have been liberated from their textual markers, suggest the possibility of other uses, other forms of circulation that exceed the repressive demands of the state.

 

[2] Self-representations

Indeed, though they were used to fulfill the demands of the state, the CI 9 photographs nevertheless trouble the distinction between the honorific portrait and the identification photograph. While the former is intended for private circulation, the latter is produced in response to a state demand. The repressive or instrumental portrait renders its subject powerless. It uses the power of the camera to capture its subject for scientific or criminalizing purposes. In contrast, the honorific portrait represents its subject in light of the traditions of bourgeois portraiture. Such images are destined to be kept in family albums, or mementos passed between friends. The subjects pose with props and costumes that indicate the positions they aspire to, even if they do not necessarily occupy them in the moment the photograph was taken.

Thus, Cheng Dau wears a tailored suit and reclines in a chair as he places his hands on the hat resting on his lap. Neither the suit, nor the hat, nor his posture betrays the fact that he has laboured in Canada for a decade as a laundryman. It is only the accompanying text that contradicts the posture of bourgeois comfort he adopts in the photograph. While some of these photographs portray their subjects in plain, dark clothing, many of them feature subjects in carefully tailored Western-European-style dress. The men wear stiffly starched collars, carefully knotted ties, and suits. These are photographs of subjects who actively participated in the way they would be represented. They convey a dignity and respectability that confounds the very stereotypes that have led to their subjection to this form of state scrutiny. In this photograph, Cheng Dau turns away, quite literally, from the visually flat and symmetrical pose that is typical of identification photography. As Pegler-Gordon understands, “Chinese immigrants played a role in shaping their representations of themselves” as a means of negotiating and contesting their racial marginalization (2009, 42). While Pegler-Gordon’s work is focused on the archives of Chinese immigration to the United States, her suggestion that variations in identity photographs constitute a form of contestation and resistance remains pertinent to Chinese-Canadian identity photographs.

The CI 9 photographs illustrate the way in which Chinese immigrants had a strong role in shaping how they would be identified by the state. As the photography historian Roberta Wue observes, “as much as the photographer may have posed and directed each photograph, each client inevitably brought his or her own preferences and expectations to the making of the portrait.” Posing for the camera for the CI 9 photographs, Chinese immigrants take a repressive state measure and transform it through an existing honorific visual tradition. As Strassler points out, state-imposed identification measures continually “rub up against competing local, familial, and religious forms of identity and sources of recognition” (2010, 129). The line between the honorific and the repressive remains blurred in the CI 9 photographs.

Formally, even though most of these head tax photographs can be located within the protocols of repressive photography, these images are also entirely self-aware and consciously styled portraits. It is clear from the use of backdrops, props, lighting, and the oval shapes of some of the prints that these photographs were taken in a professional photographer’s studio. This commercial backdrop is itself an indication of the connection between instrumental and bourgeois photography manifested in these photographs. The backdrops and props strongly suggest that these portraits circulated further than an identification photograph on a government document. The images show that there are modes of responding to the demands of the state that transform this repression into an occasion for self-representation.

The arguments for a more agential approach to identification photography made by Pegler-Gordon and Strassler mark a departure from the Foucauldian protocols for reading portraiture developed by Tagg. Nonetheless, Pegler-Gordon does allow that there are significant differences between honorific portraits taken as a form of self-representation and repressive photography taken “for the purposes of criminal identification or scientific observation” (2009, 42). As Tagg suggests, honorific portraits often include identifying props, soft lighting, and what he calls “cultivated asymmetries” (1988, 36). In contrast, repressive photography strips its subject of distinguishing props, uses harsh lighting, and places its subject in the centre of the frame. These instances of repressive photography follow in the tradition of the mug shot as delineated by critics such as Finn. However, as Sekula notes of modernist photography in the works of photographers like August Sanders and Walker Evans, the model of the archive – for example, the police photographs that inspired Evans – informs the work of art photography. Indeed, it is not at all coincidental that images of racialized subjects reveal a moment prior to the polarization of affect between honorific and repressive portraits.

Further, the connection between a cultivated asymmetry and the honorific does not hold when we turn to Chinese traditions of portraiture.

 

Chinese Traditions of Portraiture

Instead of assuming that the Chinese migrants who posed for these identification photographs were contorting themselves in order to conform to a repressive demand based on Western European modes of representation, it is also possible to read the CI 9 photographs as taking part in, and injecting into the archive of Western repressive photography, non-Western modes of photography, specifically portraiture. As Yi Gu recognizes, “By the time the Chinese encountered photography, portraiture had already reached an expanded and democratized market” (2013, 121). Relatedly, Roberta Wue demonstrates that Chinese portrait traditions have a long history of precisely the same kind of visual format associated with standard identification photographs. Early photography in China was dominated by portraiture. “After its arrival in Hong Kong in the 1840s,” Wue notes, “photography was almost primarily used to make portraits” (2005, 262). She points to the close relationship between traditions of Chinese portraiture painting, particularly ancestor portraits, which have been in use since the Song Dynasty (960–1279), and early Chinese photographic practices.

As Wue recognizes, there is an emphasis on flatness in both Chinese painting and photographic portraits: “In ancestor portraits, the face is nearly reduced to its constituent components, with every feature emphatically and clearly represented, including both ears, and with profile and partial views of the face avoided. Something of this approach to likeness carried over into photography. In fact, photographers even powdered the sitter’s face or later retouched the face in the photograph in order to eliminate unwanted shadows or marks” (269). Chinese portraiture’s emphasis upon uniformity and visual flatness went directly against “two assumptions of nineteenth-century Western portraiture in general and American portraiture in particular: that the individual and his or her inner life could be understood through telling physical idiosyncrasies and that conveying the sitter’s individuality was crucial to a successful portrait” (261). In contrast, Chinese demands for portraiture emphasized a completely different set of visual desires. Wu Hung reports that, around the late 1860s and 1870s, “the Chinese ‘peculiarities’ in portrait taking also became a recurrent topic of ridicule in Western accounts: a local sitter always demands a full frontal view with both ears showing, always looks straight into the camera lens in a confrontational manner … always hates shadows on his face, always wears his best clothes” (2010, 80). Understanding the visual history of portraiture in China allows for a recasting of the identification photographs in the CI 9 archive where the very elements that make these photographs seem like superlative examples of repressive photography are the same elements that Chinese portrait sitters sought out.

I am not the first to see some connection between identification photographs and ancestor portraits. Jan Stuart writes, “The ancestor portrait is reductionist – much like an unflattering modern ID photograph that aims to reproduce key features of a face clearly, without trying to arouse an emotional response from the viewer. Western artists trying to balance the desire to capture a key kernel of personality while also projecting a faithful likeness of a person often turn to a three-quarter view as the most satisfactory, but that view was unacceptable in Chinese ancestor portrait with its desire to hide nothing in the face” (2005, 224). Stuart rejects a Foucauldian approach to the genre and argues instead that interpreting the subjects of these images “as effigies that fulfill socially constructed roles” results in a failure to see the ways in which these images produce “fidelity to personal appearance” that results in a “misrepresent[ion of] the genre” (201). Further extending her analysis to the ways in which these images might be consumed by non-Chinese viewers, Stuart suggests that the West’s interest in portraits as offering psychological insight into their subjects actually distorts the ability of Westerners to portray and see these images as physically accurate in their representations of their subjects:

The tendency of foreigners to read Chinese ancestor likeness as homogenous images may stem, at least in part, from the unconscious application of standards used in the West to judge portraits since the Renaissance. From the late fifteenth century in Europe, so-called humanistic portraits that reveal a nugget of the sitter’s personality have been appraised as naturalistic regardless of the manipulations the painter indulged in to lay bare the soul of the sitter. This desire to privilege psychological insight over clinically accurate and emotionally detached description of physical features reveals priorities diametrically opposed to those underlying Chinese ancestor portraits. (ibid.)

Stuart’s claim indicates a way in which the supposed flattening of features found in Chinese ancestor portraits might actually be more accurate, more attuned to representing the individual, than supposedly natural or humanistic portraits emerging out of the European Renaissance tradition.

The Western emphasis on individuality, and the faith in photographic portraiture as a direct line to that individuality, runs counter to a different understanding of portraiture in Chinese photography studios. Discussing an anonymous report in an 1875 edition of Anthony’s Photographic Bulletin on Chinese photography, Yi Gu notes that the report is “laden with racial bias,” but it also records a common phenomenon: photography studios without cameras. Instead, the observer reports, “You could go in at any hour of the day, providing you were a Chinaman, and get a portrait executed in a very short time … The heathen Chinese had merely acquired a large collection of portrait negatives, and when a customer came, he took his measure mentally, looked through all the stock, and chose the picture most like. As all Chinese heads are pretty similar, and their pigtails much about the same length, it was never difficult, apparently, to make a match, for the public were quite content with what they got for their money” (2013, 120). As Gu recognizes, the story is ostensibly humorous because of “the sharp contrast between photography’s supposed faithful reference to reality and Chinese clients’ alleged acceptance of generic images as their own likeness. ‘All Chinese heads are pretty similar’ echoes the contemporaneous Western complaints about the formulaic nature of Chinese portraiture” (ibid.). More than just an indication of Chinese approaches to portraiture that pull away from naturalism and realism, in the context of identification photography, this anecdote also reveals a different way of relating to indexicality. In addition to the racism of the anecdote, let us rest for a moment on what it reports: at least as early as 1875, Chinese people frequently went to photography studios in China, and, after perusing a number of pre-existing portraits, would select the one they felt most resembled themselves, and purchase copies of that image to be used as a self-portrait. This practice of using a generic image as a self-portrait points at an entirely different approach to identification photography. Here, there is a pre-existing practice of portraiture that claims verisimilitude through broadly defined forms of likeness. The body in the photograph does not have to be the body that claims that photograph as its own likeness.

Look, for example, at an exceptional instance of this practice in the CI 9 archive. Here, there are two bodies posed together to represent one. The body in either photograph does not have to be the body that claims to be represented on the certificate. But, as the multiple signatures and counter-signatures make clear, both were accepted as representations of the person attached to the certificate.

Departing from Victoria on the Bellerophon on 1 November 1911, alongside the migrants I discuss earlier in this chapter, Ma Chow Lai’s CI 9 contains not one but two identification photographs. Look carefully at them both. Although they share some generic likeness – both of the photographs depict a person with little hair at the crown even though it is unclear if this person is balding or has fashioned their hair into a queue – it is not at all obvious that they are images of the same person. The shape of the face is distinctly different. The lips do not seem to be the same. The eyes and eyebrows also appear to be quite different from one image to the next. And yet, at one point, both photographs were accepted as representing the same person. Here, a generic likeness suffices not once, but twice, as an identification photograph. Not only did Ma Chow Lai, or the person or people claiming to be Ma Chow Lai, insist upon the correctness of both portraits by deploying them on the CI 9, he (or they) also convinced the immigration agent to accept these versions of likeness. In an echo of the custom of choosing a likeness from a pre-existing selection outlined by Yi Gu, Ma Chow Lai’s doubled portrait poses a challenge to the outlandish idea that one image should suffice as an index to a single identity.

How this need for both photographs arises on this particular certificate remains a mystery. We have only the evidence before us: the material on the certificate itself. According to the certificate, Ma Chow Lai returned to Victoria less than a year after departing and was admitted back into Canada on 30 August 1912. Did the admitting agent check for the facial marks and peculiarities noted on the form? Did the man who returned in 1912 also have a scar and pits on his forehead, a scar on the top of his head, and a scar on the left side of the bridge of his nose? These marks are not visible in either photograph.

One photograph, the one on top, is pasted over the bottom one. It is possible that the bottom one was original to the certificate, and the top one was appended as a form of further verification. In this deployment of a doubled image, CI 9 no. 18352 is unique. I have not come across any others that contain two photographs purporting to identify one person. In its exceptionality, it both reveals a number of singularities specific to this certificate, and also illuminates the obverse, the practices and norms that regulate the vast majority of the other certificates.

Not least, its singularity suggests that the majority of identification photographs were taken in Canada. Scribbled in the lower left-hand corner of the certificate are the words: “Photograph taken in Hong Kong.” There are no other notes about the origins of the photographs in any of the CI 9s that I have encountered. How a photograph originating in Hong Kong might explain the need for two distinct photographs is not at all clear. Perhaps it was an explanation for the change in appearance between the man who departed and the one who returned. Perhaps it suggests that the first photograph was inadequate. Whatever the explanation, that note in the corner does imply that it was unusual for a photograph to be taken in Hong Kong and used on a CI 9.

This doubled identification photograph also shows the way in which a single image was almost always expected to serve as a sufficient tool of identification. Indeed, not unlike our contemporary passport photographs, the CI 9 system’s reliance on one photograph does not account for how a person might change over the passage of time. As I will discuss in the next chapter, even where there are photographs of children who return many years later, and thus presumably would have changed a great deal, the single original image of the child is still sufficient proof of identity for readmittance.

Ma Chow Lai’s Hong Kong photograph is the most visible evidence of the way in which Chinese photographic and portraiture practices and customs became threaded into the CI 9 system. Even though this system was put into place to capture Chinese migrants on a massive scale for the purposes of excluding them from citizenship, the system did not merely impose a Western European mode of surveillance upon its Chinese subjects. The system was also susceptible to the ways in which Chinese migrants imposed on it a different set of codes and values. Not only was the CI 9 system ill-equipped to remember those it captured on the basis of their given names, it also possessed no bulwark against different ways of seeing and being seen that Chinese migrants inserted into each certificate with each photograph.

I want to be cautious and resist claiming that there is an easily distinguishable and homogenous Chinese tradition to which we can ascribe so much of the visuality of the identification photographs on the CI 9 certificates. Of course, China was and is a deeply diverse site of photographic practices, and there is no single, continuous, uncontested photographic tradition. What is more, I take seriously Wu Hung’s cautions against the almost too-neat dichotomies of Western and non-Western visualities that can be drawn from these contrasts between Chinese and Western portraiture. Or, as Gu notes, the story of the photography studio without cameras “neatly embodies the dichotomies – Western versus Chinese, naturalistic versus fanciful, realistic versus formulaic – that beset Chinese visual culture after its head-on encounter with Western imagery, ideas, and techniques” (2013, 120). Wu Hung challenges the idea that there was a distinct divide between Chinese and Western portrait styles:

As Régine Thiriez summarizes, after having scrutinized hundreds of old portraits from the nineteenth and twentieth century, generally speaking there are no radical differences between products of Chinese and Western studios. Chinese studios adopted Western formulas in setting up the backdrop and arranging the sitters’ poses, whereas many works by Western studios nicely fit the aforementioned Chinese conventions. Her conclusion is supported by Cody, Terpak, and Fraser in their catalog essays and by my own study of photo archives in China and abroad. Rather than confirming the sharp divide reported in the Western accounts, an empirical study of primary data reveals instead an abrupt split between practice and discourse. (2010, 81)

This skepticism over any kind of intrinsic East-West aesthetic divide is warranted. At the same time, though, it is absolutely the case that Chinese portraiture has a long and robust tradition that did influence the rise of photography in China. As Gu notes, “While we may downplay the emphasis on ‘Chinese peculiarity’ as a mere reflection of colonial anxiety, traces of stylistic distinction were manifested in a good many photographs of Chinese sitters” (2013, 122). Gu attempts to recognize the specificity of Chinese photography, without falling into an easy East-West binary, through a rigorous investigation of the Chinese names for photography, their evolution, and the implications of these names for illuminating an approach to photography that could be identified as specifically Chinese.

Through her investigation, Gu draws attention to the intimate relationship between painting and photography in China. “The first names for photography in Chinese – yingxiang, xiaoxiang, xiaozhao – were all preexisting terms for portrait painting” (2013, 121). For Gu, China’s encounter with photography needs to be understood as part of a broader complex of visual practices where painting and photography are not diametrically opposed but, in fact, synchronous. “If there are antinaturalistic traits in Chinese photography, they neither indicate a lack of understanding by the Chinese of photography’s media specificity nor demonstrate a conscious resistance to it. The fact that the names first used for photography were all preexisting words for portrait painting highlights a historical moment in China when photography belonged to a rapidly changing and expanding field of visual practice that was conveniently dubbed ‘painting’” (ibid.). Gu’s interventions break down the divide between photography and painting in Chinese visual practices. Indeed, as she shows in her discussion of photography studios in nineteenth-century China and in her engagement with the work of early photography critics such as Wang Tao and Du Jiutian, from the inception of photography in China, there was an intimate connection between painting and photography.

While Gu’s analysis focuses on Chinese photographers in nineteenth-century China, Catherine Clement’s Chinatown through a Wide Lens: The Hidden Photographs of Yucho Chow shines a light on how photographic portraiture – and its connection to painting – was practised in early twentieth-century Canada. Yucho Chow operated a photographic studio in Vancouver’s Chinatown from 1907–49. As Clement notes, Chow operated this studio during a period of intense change and turmoil for Chinese Canadians: “He started photographing during the time of escalating head taxes and the famous 1907 Race Riot. His camera operated through the First World War; the 1923 Chinese Exclusion Act [the Chinese Immigration Act was amended in 1923 to exclude virtually all Chinese immigration until its repeal in 1947]; the Great Depression; the Second World War; and finally, the repeal of the Chinese Exclusion Act and the granting of voting rights in 1947 to several communities, including the Chinese” (2019, 25). In this mapping of the historical period of Chow’s studio, Clement draws attention to the connection between Chow’s photographic work and Chinese Canadians’ trajectory from exclusion to citizenship. Chow’s studio focused almost entirely on portraiture. Aside from a handful of images of Chinese businesses and Sikh temples, the archive of Chow’s photographs is overwhelmingly made up of portraits of people in the Vancouver community: “Over the course of four decades, he captures photos of people across the arc of life’s journey: from newborn babies to the recently deceased. (These ‘death’ photos were often sent back to family in China and Europe and served as a form of death certificate for grieving relatives)” (33). His studio moved at least four times within Chinatown, but it was, as Clement points out, one of the very few places that welcomed people of all races and classes. Thus, Chow captured portraits not only of Chinese Canadians, but also those from South Asian and European communities. Although the majority of his images were taken in his studio, Chow would also “haul his heavy equipment down to Vancouver’s central train station where he regularly took head and shoulders shots of bleary-eyed, newly-arrived European immigrants. Ironically, these were the very immigrant groups that the federal government was inviting to Canada even as his own community was increasingly stigmatized and unwelcome” (ibid.). Chow’s studio portraits involved carefully chosen props to connote wealth and status, such as area rugs, distinctive backdrops, faux marble pedestals, intricately carved furniture, and porcelain vases.

The connection between painting and portraiture in Yucho Chow’s portraits can be found not only in the composition of the portraits, but also in his backdrops and use of hand colouring. His portraits were distinctive for the backdrops he constructed and painted himself. Practised photographic historians such as Clement can quickly spot a Yucho Chow portrait by looking at the backdrop alone. Chow’s portraits were also particularly notable for their hand-painted embellishments. Chow’s daughter Jessie (1916–1999) “learned to hand paint the photos using oils. She became so skilled that some photographs look as if they were originally shot on colour film” (Clement 41). While many photographers, whether they are Chinese or not, deploy some combination of all of these painterly practices (compositions, backdrops, hand colouring), Chow’s use of them in this early period of Canadian photographic portraiture offers a bridge between the period of Chinese portraiture that Yi Gu examines, and the CI 9s in my analysis. Although there is no clear indication that Yucho Chow took photographs for use on CI 9 identification photographs, his focus on portraiture indicates that a Chinese-Canadian photographer from this period did indeed bring knowledge of Chinese portraiture traditions into early Canadian photographic practice.

As Yi Gu argues, for Chinese photographers, the distinction between painting and photography was mutable. One was not opposed to the other and they shared practices and techniques. These visual traditions not only break down the binaries between photography and painting and between East and West, but they also show how the migrants in the CI 9 photographs might have used these photographs to mediate a refusal of the protocols of identification photography where one body was to be so insistently indexed to one image affixed to one certificate. Instead, these photographs attest to a range of refusals: to be identified; to be fixed; to be cut off from honour and dignity.

Even though the stipulations of identification photographs might read as repressive, they are also evocative of an alternative visual history of Chinese portraiture. Viewed from within a continuum of visual practices that connect photography and painting, the CI 9 photographs offer a representational through line connecting ancestor portraits and identification photography. The very characteristics that indicate the repressive measures of the state – visual flatness, full-frontal exposure of the face, minimization of light and shadow – are also the characteristics of successful, honorific portraiture in China.

Punched through, filed, and stored, the CI 9s are a testament of the failure of memory when it is conflated with storage. They are a tribute to the multiple possibilities for evasion and escape. Hiding while still inhabiting the technology of mass capture, Chinese migrants manufactured forms of inclusion for themselves that far exceeded the limits of non-citizenship.