Composite
The word composite stems from the Latin word compositus, which means “to put together” (“Etymology of Composite”). The word composite is a composite itself; the prefix “com” means “together,” while posit means “to place.” In this sense, composite’s definition encompasses its surface-level function: the assembly of parts to create a cohesive whole. Galton and Daston highlight the overlap between composite photography and the composite human that enables the highest degree of political utilitarianism. Du Bois takes a different route, choosing to use statistics, rather than a camera, to create a composite of a race. With this, we can identify how these different approaches ironically combine to enhance the word composite.
The composite Daston outlines in her “image of objectivity” has two layers: the physical components of a composite photograph and the composite of objectivity and technicality used to produce the image (Daston 81). An examination of Galton’s synthetic criminal can be used to illustrate this notion. Galton employs advanced photographic technology that superimposes details from other criminal mugshots onto a single canvas. He does this in order to eliminate subjective impulses and uncover facial features indicative of criminality (Daston 102). Galton’s work, in this sense, contributed to a larger trend of quantifying human attributes through the use of composites in order to limit psychological biases. In this sense, composite becomes a term that blends disciplines such as psychology and statistics, in addition to physical elements. This claim is complicated through W.E.B. Du Bois’ composite data portraits that illustrate the post Civil War progress of African Americans (Du Bois 12). Du Bois creates a composite through a cross-reference of his Georgia and diasporic collections, using a star to signify Georgia’s presence in the African trade route (Du Bois 11). With this, composite’s meaning is further developed. Du Bois’ model backs Galton’s initial usage of composite to eliminate subjectivity and plant roots in statistical models. Yet, most notably, he takes the concept of a composite a step further in order to produce a narrative. This, therefore, creates a paradox. Du Bois’ collections are intended to prompt observers to change their previous perception of African Americans. In order to do this, however, he uses composite statistics to paint stories, such as “representations of Harlem in the twentieth century” that seek to disprove racial misconceptions (Du Bois 13). With this, Du Bois’ work builds off of Galton’s in order to highlight the potential reverse effects of using a composite to eliminate human interference. In reality, the usage of specific statistics may retain a degree of human interference within composite models.
We can apply composite’s roots in objectivity and its potential perverse effects to Galton’s anthropometric apparatus. Galton’s model stresses the use of anthropological data and photography in order to create the ideal composite human. This “objective” data could theoretically be used to create a superior population, devoid of all physical inadequacies (Galton). For example, Galton’s ideal composite human would have the statistically-optimal weight, hand strength, reaction time, and other physical characteristics necessary to maximize utility (Galton 4). Yet, Galton’s motivations are troubling towards the principle of objectivity. He claims there must be a typical form of human towards which the human race should “chiefly cluster” (Galton 10). In an attempt to quantify human features with unbiased statistics, therefore, Galton furthers a parallel, subjective motive, similar to that of Du Bois. This falls into the previously-uncovered paradox and prompts an examination of composite usage within humans. In this way, Galton’s composite confuses biased data with reality, proving that composites are not objective tools in the context of anthropology.
Galton’s work can be seen as an instrumentalization of Baudelaire’s anxiety towards the fraction of the aura. Baudelaire claims that as photographs are altered, their aura is consequently fragmented and diminished (Baudelaire 83). When this criteria is applied to Galton’s composite images, it assumes the original photo’s aura is severely compromised after so many alterations. Composite images, therefore, effectively rob humans of distinctness and individuality in their pursuit of statistical continuity. In his pursuit to create a superior human void of “subjective impulses,” Galton’s composite compromises the very essence of human existence (Galton 102). Despite composite’s etymological definition of using parts to construct a whole, Galton and Baudeliare’s work fuse to illustrate how a composite can unravel its parts in the same way it places them together. Yet, at the same time, Galton’s creations represent the birth of new art, with each piece in possession of its own, distinctive aura. Although unorthodox, even Galton’s fictional prisoner composite possesses a new aura as a result of its creation. This suggests that a composite can destroy and construct human aura simultaneously. Composite enriches the tension between loss and creation, between the individual and the components that make him. It is this dual-nature that blurs the line between Galton and Baudelaire’s logical parameters.
Galton’s plight to construct the ideal composite human reflects his motivation to maximize social utility. He identifies his primary objective of maximizing “health,” while mitigating “disease and criminality” (Galton 9). With such parameters established, he gives rise to utilitarianism, a model that assumes actions are correct if performed on the basis of societal betterment. As a result, society becomes a composite, where each member is “layered” on one another in order to create a successful whole. For example, through a composite of royal privates, Galton identifies “vigor, resolution, intelligence, and frankness” as qualities indicative of good health (Galton 10). Yet, utility in the context of society largely differs from individual to individual; not everyone has the same parameters as Galton. In this sense, there is no “one, true composite” that satiates the parameters of utilitarianism. Rather, each individual’s idea of a utility-maximizing society combines to create a composite of everyone’s ideal utilitarian vision. Foucault’s governmentality model attempts to satiate this issue through the facilitation of utility-maximizing institutions (Foucault 93). Still, societies often fall short of complete utilitarianism. In this sense, a composite’s layers may not build upon each other evenly within a social model.
Rules
The word rules is derived from the latin word regula, which means straight stick, bar or ruler (“Etymology of Rule”). The term consequently evolved into Old French as “reule,” meaning regulation or law. In a vacuum, a rule would represent the end-all, be-all – there would be no room for further interpretation and thus no need for enforcement. Daston’s discussion of the realms of rules is complicated by Galton’s anthropometric apparatus. Within these realms, Ferguson and Barthes detail the reciprocal way in which rules are shaped, while Striphas takes these parameters and incorporates it into Babbage and Wittgenstein’s algorithms.
According to Daston, rules cover two realms: natural and human (Daston 151). A natural rule would state summer days are typically warm, while a human rule would outline the procedure for dividing an inheritance among heirs (Daston 151). With this in mind, a natural rule might discern the physical makeup of a human. As a rule, a single human should have a single head measurement, arm span, and other physical attributes that make up Galton’s anthropometric apparatus (Galton 4). Yet, Galton’s pursuit of a composite human thus violates this natural rule. In his prisoner portraits, for example, each depiction sports attributes from multiple men (Daston 102). The creation of such a prisoner could never occur in nature; yet, Galton made the prisoner appear by means of new rules within the human realm. In this sense, Galton bends the rules Daston puts forth in order to pave the way for the mechanicalization of rules. Galton’s quantification of human properties, such as keenness of hearing, play with the bounds previously put forth by the natural realm. He illustrates how the mechanization of rules can lead to their consequent departure from Daston’s two finite realms and instead bridge aspects of them together. Mechanical rules, therefore, grant humans a “divine” influence unbeknownst to Daston.
Ferguson’s cookbook builds off of Daston’s concept of human rules. His analogy of a cookbook as a democratic institution explores how rules are not always commanded, but still play an instrumental part in human development (Ferguson 1). According to this model, readers “play an active role in cookbooks, which intensifies their political force” (Ferguson 79). Through an acknowledgement of the creative and interactive aspect of rules, Ferguson demonstrates how rules are not merely restrictive or prescriptive. “Cookbooks do not transform their readers into excellent cooks” (Ferguson 14). Instead, they invite engagement, modification, and imagination. Such interaction is inherently democratic – individuals are empowered to improve and adapt upon rules in the same way democracy encourages an evolutionary perspective on its principles. Just as a recipe might be adapted to suit taste preferences, rules on same-sex marriage might change to fit altered public preferences. On the flip side, rules such as the second amendment exist as a concession to the fact that we need not be devoted to a set of rules. This excuse parallels ideas Roland Barthes puts forth in “Death of the Author.” As the author’s aura is removed from the text, the reader is able to interpret it through his own eyes (Barthes 142). In this sense, the author’s rules of interpretation are removed, enabling the reader to approach the text democratically. For this reason, the absence of or alleviation from rules can be just as powerful as their institution.
The democratic nature of rules put forth by Ferguson paves the way for its incorporation into algorithms. Striphas defines algorithmic culture as “the expanding methodological ambit of key-words to include mathematical units, concepts, operations, and more,” largely by means of quantitative metrics (Striphas 72). Culture and computation are thus blended into a set of rules beyond human visibility. For example, Striphas’ describes a couple that is assigned a high probability of fraud by an insurance algorithm and consequently denied a claim after a horrific accident (Striphas 68). An algorithm made “fundamental judgments about who gets to count as important in [their] lives, how, to what extent, and in what specific contexts (Striphas 68). In this sense, rules are no longer clear-cut and are thus forked into two realities. The first is Babbage’s “Difference Engine,” which would attribute the insurance algorithm’s conclusion to the “mechanical gearwork” of the machine (Daston 142). Here, there are no exceptions to the rules – the algorithm’s anomaly can be attributed to “apparent miracles inscribed in the machinery of nature by the divine engineer” (Daston 142). This, in a sense, merges Daston’s human and natural realms. In the same way nature operates based on rules from a divine creator, the “divine engineer” designates an algorithm as the natural vehicle for the diffusion of rules to the human realm. Therefore, under this model, rules are only enhanced with mechanical augmentation and do not compromise the two realms. The second reality lies in Wittgenstein’s version. Wittgenstein would also agree the insurance algorithm did not produce a mistake. However, he claims that such anomalies prove “the impossibility of mechanical rule-following” (Daston 142). In this model, rules would theoretically have an infinite amount of interpretations – the only way to limit this regression would be to understand the rule in the context of a tradition or custom (Daston 142). Rules, therefore, become untransferable between the natural and human realms if interpreted mechanically. The human realm produces ambiguous customs and traditions that limit the functionality of algorithms that desire clean-cut rules. Such functionality cannot be derived from “divine engineering.” The work of Babbage and Wittgenstein, therefore, illustrates how the mechanization of rules can enable or discourage their dissemination between the two realms.
Algorithm
Per Tarleton Gillespie, “to chase the etymology of the word [algorithm] is to chase a ghost” (Gillespie 18). The word algorithm has roots in the Medieval Latin word algorismus. This word, in turn, stems from Arabic mathematician al-Khwarizimi, the mathematician who introduced sophisticated mathematics to the Western world (“Etymology of Algorithm”). Gillespie’s explanation of algorithm as a synecdoche is exemplified by Galton, while Ferguson and Foucault contrast to illustrate its modes of deployment on a political scale. Algorithm can also be explained as a talisman through Benjamin’s aura.
We can overlay Gillespie’s description of an algorithm as a synecdoche with Galton’s composite in order to understand the “ill-defined network of actions upon actions” that define it. According to Gillespie, the term algorithm ignores the “hundreds of hands” involved in its production (Gillespie 22). Similarly, Galton quite literally uses hundreds of hands to create an algorithm of his own: the anthropometric apparatus (Galton 4). This conflation highlights the word algorithm’s double-edged meaning: on one hand, it hides the word’s sometimes unnecessary technical background, but on the other, it leaves room for political values to slip in (Gillespie). Likewise, we can examine instances in which Galton lets personalized parameters infiltrate his algorithm. For example, those with vision and hearing disabilities would not satisfy vision and hearing keenness tests outlined in Galton’s algorithm (Galton 8). This decision is controversial; many might feel disabled persons have valuable qualities to offer to Galton’s ideal human. Yet, Galton is able to sweep this detail under the rug, using the grandeur of his anthropological algorithm and its promotion of social perfection to mask it. In this sense, the technical aspect of the word algorithm can consequently erase “the people involved, downplay their role, and distance them from accountability” (Gillespie 23).
As Galton and Gillespie demonstrate algorithm’s tendency to mask biases, Ferguson and Foucault examine such effects on a diverse political scale. Ferguson’s political cookbook is analogous to an algorithm in the sense that it provides a set of instructions to produce an output. With this, it is worth noting that “cookbook authors cannot tell their readers how to experience and use their texts,” thus putting “the reader, not the author in charge” (Ferguson 8). If we apply this analogy to an algorithmic model, it suggests that the user, rather than the developer of the algorithm, is in charge. For example, a user may vary inputs they provide to an algorithm, which is akin to altering ingredients in a recipe based on personal preferences (Ferguson 7). An algorithm might also use user feedback to reshape its accuracy, just as cookbook authors seek to include sought after dishes. Therefore, in the same way one reads a cookbook and ends up contributing towards it, users of an algorithm contribute to its composition. Thus, algorithms are inherently democratic through this medium. This phenomenon contrasts with Foucault’s governmentality model. If we equate an algorithm to a government under his model, its sole purpose is to preserve the “political economy” and “apparatuses of security” (Foucault 102) in order to “lead to a convenient end” (Foucault 93). In this model, a government’s political algorithm lies not in the advancement of culture and traditions that allow people to shape it, but in the facilitation of institutions that shape them. For example, Ferguson’s governmental algorithm would encourage its people to form a new law (or dish) through the diffusion of cultures and traditions, while Foucault’s would enable the same process through educational institutions that promote such ideas (Foucault 102). With this, algorithms allow us to “rethink how authority, commands, and directions operate (Ferguson 6). Thus, the two approaches identify two ways in which governments exhibit different algorithmic approaches towards their subjects.
The political nature of algorithms is threatened in its entirety by the more rudimentary ideas of Wittgenstein and supplemented through a study conducted by Malthus. He claims neither humans nor machines can follow rules in a purely mechanical way, including through an algorithmic medium (Daston 142). With this in mind, if algorithms exist as an abstract set of rules, then they truly are inhuman, and thus apolitical. Instead, they create a system in which reality is merely a rendered output or correlation of data. For example, Malthus details a study in which the birth to burial ratio was 124 to 100 in one period, but unexplainably 111 to 100 in another period (Malthus 39). Malthus’ statistical algorithm failed to acknowledge that “in the natural progress of the population of any country, more good land will, caeteris paribus, be taken into cultivation in the earlier stages of it than in the later” (Malthus 39). In this sense, an algorithm existed only in a data-driven reality and failed to account for a variable rooted in human behavioral trends. This emboldens Wittgenstein’s claim that algorithms have mechanical limitations and consequently highlights their inefficiency in people-driven political models.
A final dimension of the term algorithm is its use as a talisman, oftentimes to “ward off criticism” (Gillespie 24). For example, if a company encounters an error, it can be attributed to “the algorithm” (Gillespie 24). This, in turn, suggests an algorithm is beyond the average person’s understanding, thus imbuing it with a magical quality. It also reflects how people may view an algorithm as infallible, even when that is not the case. This ties in to Benjamin’s concept of the aura, where an algorithm develops a “strange tissue or space and time” that draws the masses towards it (Benjamin 23). This represents an interesting duality. Previously, an algorithm’s focus on qualitative, data-driven metrics drew it away from human culture and tradition. However, if an algorithm possesses an aura, it might create a culture around itself, therefore infiltrating human nature in its own way. In this way, algorithm as a talisman allows it to cross into the human realm.
Works Cited
Baudelaire, “The Modern Public and Photography” in Alan Trachtenberg and Amy R. W. Meyers, Classic Essays on Photography (New Haven, Conn: Leete’s Island Books, 1980)
Barthes: “The Death of the Author” in Image-Music-Text. Macmillan, 1978.
Benjamin, Walter. “The Work of Art in the Age of Its Technological Reproducibility: Second Version.” In The Work of Art in the Age of Its Technological Reproducibility, and Other Writings on Media, First Edition., 19–56. Cambridge, Mass: Belknap Press, 2008
Daston, Lorraine. Rules: A Short History of What We Live By. Princeton University Press, 2022.
Daston, Lorraine, and Peter Galison. “The Image of Objectivity.” Representations, no. 40, 1992, pp. 81–128. JSTOR, https://doi.org/10.2307/2928741Links to an external site..
Du Bois, W. E. B. W. E. B. Du Bois’s Data Portraits: Visualizing Black America. Edited by Whitney Battle-Baptiste and Britt Rusert, Princeton Architectural Press, 2018.
“Etymology of Algorithm.” Etymology, www.etymonline.com/word/algorithm. Accessed 5 Mar. 2024.
“Etymology of Composite.” Etymology, www.etymonline.com/word/composite. Accessed 5 Mar. 2024.
“Etymology of Rule.” Etymology, www.etymonline.com/word/rule. Accessed 5 Mar. 2024.
Ferguson, Kennan. Cookbook Politics. University of Pennsylvania Press, 2020.
Foucault, Michel. Discipline & Punish: The Birth of the Prison. Knopf Doubleday Publishing Group, 2012
Galton, Francis. Composite Portraits. Harrison and Sons, Printers in Ordinary to Her Majesty, 1878.
Galton, Francis. Generic Images. Royal Institution of Great Britain, 1879.
Galton, Francis. Inquiries into Human Faculty and Development. Thoemmes, 1998.
Galton, Francis. The Anthropometric Apparatus. The Cambridge Scientific Instrument Company, 1887.
Gillespie, Tarleton. “Algorithm.” In Digital Keywords: A Vocabulary of Information Society and Culture, edited by Benjamin Peters. Princeton University Press, 2016: 18-30.
Gilman, Sander L. Review of A Life of Sir Francis Galton: From African Exploration to the Birth of Eugenics. Perspectives in Biology and Medicine, vol. 45 no. 3, 2002, p. 468-470. Project MUSE, https://doi.org/10.1353/pbm.2002.0046.
Malthus, Thomas Robert. An Essay on the Principle of Population. Cambridge Univ. Pr., 1989.
Striphas, Ted. Algorithmic Culture Before the Internet. Columbia University Press, 2023. Read: Ch. 2, Algorithm