'First-Class Objects' by James Grimmelmann in (2011) 9
Journal of Telecommunications and High Technology Law 421
asks
What is the difference between "James Grimmelmann" and "@grimmelm" and why should we care? Some computer systems, like Facebook and credit reporting agencies, are inherently "about" people. Others are not. This essay argues that the key technical difference is whether they use unique identifiers to refer to people in their databases. From this single distinction, a host of social and humanistic consequences follow. The essay taxonomizes them and teases out some of their implications for privacy law.
He goes on to state that -
How should we think about privacy in a digital age? One approach is to focus on how people use computers: how what we choose to share about ourselves changes when we go online. But we could also focus on how computers use people: how flows of personal information are transformed by technology. Just as email is different from mail, a spycam is different from a spy.
This brief essay will examine a seemingly technical question: how are people represented within computer systems? The essay will argue that that there are two possible ways to do it, and that the choice between them has important technical, social, and humanistic consequences. It won’t say much new about those consequences — instead, it will show how closely linked they are. ...
the transition from unstructured data to structured data is of critical importance for thinking about privacy and social interactions. There are echoes of at least three previous shifts in this transition: the introduction of print, the growth of bureaucracy, and the rise of digital media. All three of them have reworked the relationships of individuals to each other, and to the larger institutions that make up their worlds: communities, companies, and countries. The use of unique identifiers as the keys to structured databases about people will have its own dramatic consequences.
Another computer science term, this one from the field of programming languages, is suggestive of the values at stake. One sometimes says that a system which directly represents certain things treats them as “first-class objects.” One computing website explains that an element in a programming language is first-class “when there are no restrictions on how it can be created and used.”
For example, in some programming languages, like C, functions are not first-class. Any subcomputation that the program will carry out must be specified in advance by the programmer, and there are significant limits on how functions can be stored, modified, and passed around. In other programming languages, like Scheme, functions are first-class: the computer treats them just like it would any other kind of data, like a number or a binary true/false. This leads to great flexibility. Scheme programmers can add new functionality on the fly as the program runs; they can do clever things with functions that C programmers can only mimic imperfectly and at much greater length. It is easier to work with and reason about functions in Scheme than in C, because functions are first-class in Scheme and not in C.
People are first-class objects on Twitter: it has the capacity to distinguish and reason about them. The same is true in the many other systems that give people unique identifiers as a way of representing them in databases. Both halves of the phrase are illuminating. On the one hand, people are truly first-class: this representation enables useful features that connect directly to these individuals’ wants and needs. On the other, people are also objects: when these systems represent people, it is often without their knowledge or consent.
... treating people as first-class objects - representing them with digital identifiers - has significant technical and social consequences. Perhaps it should have legal consequences as well. We should expect the creators of these first-class objects to take care to treat people with the respect and concern the name suggests they deserve.
