https://d2ng68cf36kxsk.cloudfront.net/store/public/post/775/post_images/Slide234.JPG">

Intra-taxonomy relations are created by inclusions, exclusions, and cross-references.¹⁶⁹ The introductory text at the beginning of the standard sternly warns of the dangers of using word processing software such as Microsoft Word for storing and sharing classifications. XML rigorously represents and preserves taxonomic structures, it says, and facilitates interoperability between different medical software applications.¹⁷⁰

ICD-10 is an ontology, technically speaking in the discourse of information science, and also, we would argue, philosophically as well, framing as it does meanings in the play of the ideal and the material. XML is a way to write ontologies, or schemas as they are often called.

We use the word “write” advisedly here, because this is a peculiar genre of spatialized, visually arranged writing that again demonstrates the dramatic differences between speech and writing, particularly in the era of digital text. Once more, we find reason to abandon the category “language” for its too-easy conflation of speech and writing, as if the one were a mere transliteration of the other.

Among the many mostly unnamed inventors of our forms of everyday life in the era of digital meaning is Thomas Gruber, co-creator of Siri, the voice interface that Apple bought in 2010 from his startup of the same name for a reported price of between $150 and $250 million. Like many of the other innovations in the era of digital text, his original research had been funded by the US Department of Defense.¹⁷¹

Gruber, however, may be more significant for establishing the notion of “ontology” in the emerging infrastructures of digital representation. “An ontology is an explicit specification of a conceptualization … the objects, concepts, and other entities that are assumed to exist in some area of interest and the relationships that hold among them.”¹⁷² It is, he says, “a set of representational primitives with which to model a domain of knowledge or discourse.” These are “typically classes (or sets), attributes (or properties), and relationships (or relations among class members) … designed … to enable the modeling of knowledge about some domain, real or imagined.”¹⁷³ With the concept “ontology,” Gruber managed to crystalize a number of longstanding issues in computer science about the connections between its syntax and semantics.¹⁷⁴

In the terms we have been developing in this grammar, ontologies represent structures of meaning with digital text. An ontology purports both to represent meanings-in the material world (for instance, the health of bodies), and meanings-for the world in our ideal figuring so we can serve our meaning-filled interests (for instance, the practice of medicine).

This is an essentially textual practice, where the form of writing in the digital era separates itself further and further from the form of speech. When from version to version ICD terms increase more than fourfold and their differentials by an unquantifiable number, the possibility of speaking medicine spontaneously from memory becomes less and less feasible.

If “natural” language is the speech of everyday life, used for everyday bodily purposes such as medicine, for science and health’s sake the textual practices of medicine have become strategically unnatural. This is why medical practitioners need their IMO apps^§MS3.2a and other such digital resources on their phones or tablets. It is also why medical education needs to pay more attention to practitioners’ communication with patients because the gulf is growing between their textual practice and patients’ speech.

On the other hand, when you ask a question of an intelligent machine assistant such as Siri and she talks back, the speech part is not the core of the technology. It is the ontology part, or knowing what to say.

“Phone the nearest Indian Restaurant,” you might ask Siri. If you get an answer, it is because the computer looks up structured text. “Sitara” is not just a word, it is marked up from ontologies as a restaurant whose property is to serve Indian food. Its address in Urbana and its phone number are not just words and numbers; these are marked up as geolocations and phone numbers. Siri’s slightly stilted response is only smart because the ontologies that define structured text are smarter than what from a computer scientist’s point of view is considered unstructured text. Often, Siri remains speechless, and you just get a link to a website – this is because the data is not clearly enough marked up or not structured in a processable ontology.

Siri, the human behind the machine, is a voice actor, Susan Bennett. Working at home in Atlanta, Georgia, she took a month in 2005 to record meaningless fragments of speech for Gruber’s company. Before this she had been the voice of advertisements for McDonalds, Coca Cola, and Ford. Bennett didn’t know Apple was going to use Siri in iPhones until the application was launched in 2011. Now she’s another ubiquitous but anonymous presence in our digital lives.¹⁷⁵

---

¹⁶⁹ Stefanie Weber, Egbert van der Haring, Susanne Bröenhorst, Michael Schopen, and Pieter Zanstra, 2005, “Maintaining Medical Classifications in XML: ClaML Redefined for Use with WHO-FIC Classifications,” paper presented at the WHO-FIC Network Meeting, 16–22 October, Tokyo, Japan. http://apps.who.int/classifications/apps/icd/meetings/tokyomeeting/B_3–1%20Maintaining%20Medical%20Classifications%20in%20XML-ClaML.pdf

¹⁷⁰ Health Informatics Technical Committee CEN/TC 251, 2007, “A Syntax to Represent the Content of Medical Classification Systems: ClaML,” EN 14463:2007, Brussels, Belgium: European Committee for Standardization, pp. 5–6.

¹⁷¹ www.huffingtonpost.com/2013/01/22/siri-do-engine-apple-iphone_n_2499165.html

¹⁷² Thomas R. Gruber, 1995, “Toward Principles for the Design of Ontologies Used for Knowledge Sharing,” International Journal of Human-Computer Studies 43(5–6):907–28, p. 908.

¹⁷³ Thomas R. Gruber, 2009, “Ontology,” pp. 1963–5 in Encyclopedia of Database Systems, edited by L. Liu and M.T. Özsu, New York, NY: Springer-Verlag.

¹⁷⁴ Steven L. Lytinen, 1992, “Conceptual Dependency and Its Descendants,” Computers & Mathematics with Applications 23(2–5):51–73; Robert I. (Bob) McKay, Nguyen Xuan Hoai, Peter Alexander Whigham, Yin Shan, and Michael O’Neill, 2010, “Grammar-Based Genetic Programming,” Genetic Programming and Evolvable Machines 11:365–96.

¹⁷⁵ https://en.wikipedia.org/wiki/Susan_Bennett; www.huffingtonpost.com/2013/10/04/voice-siri_n_4043134.html

---

Cope, Bill, and Mary Kalantzis. 2020. Making Sense: Reference, Agency and Structure in a Grammar of Multimodal Meaning. Cambridge UK: Cambridge University Press, pp. 310–313. https://doi.org/10.1017/9781316459645

Thomas Gruber’s “Ontology” §MS3.3a