Notes On "The Art of Code" by Maurice Black
Notes On "The Art of Code" by Maurice Black
Notes On "The Art of Code" by Maurice Black
Black, Maurice. The Art of Code. University of Pennsylvania, Department of English, 2002. Print. [Annettes comments are in brackets. Page numbers are in the left column.] 6 He uses the Matrix to frame the intro. Hackers are the new heroes, who can manipulate the code running underneath everything we do and are in life. The message of the Matrix, which is governed by source code, and who can access it. "the victim of the future is the complacent luddite who doesn't think twice about depending on machines whose inner workings are a mystery; the savior is the hacker, the individual whose complex passion for programming forms the basis of an entire ethical system." He wants to "examine from a literary perspective an aesthetic that has always been deeply embedded in computer programming culture and that can be regarded as foundational to it." We lack "cultural histories of code" and he seeks to provide that. "abstracting code from hardware made it possible for programmers to conceptualize their work as a new kind of writing rather than as an esoteric form of mathematics. Code-sharing figures into this literary aesthetic of code very heavily, but higher-level code also allows for the "blackboxing" of code. "it is hardly an exaggeration to suggest that software's entire cultural, corporate, and legal history may be understood in terms of the interrelated effects of code's abstraction into language." Good review of software histories until then. Critique of Turkel and her artificial distinction between MS-Dos and GUI systems, and modernism & postmodernism, surface and depth. (traces her distinctions to Pirsig). "Turkleneglects to historicize the divide she posits, failing, for example, to recognize that the MS-DOS command-line interface--which arrived on the computing scene in 1981--is itself a 'scintillating surface,' the result of three decades' worth of accumulated abstraction from the real 'guts' of computing hardware. He splits the theoretical from the practical strands of computing history. Sharing of code (Hopper quote, from Wexelblat 8), and IBM's SHARE program (1955, for the 704), which contributes to IBM's project. Feenberg applies Latour to look at embedded decisions in technology. Black traces this to
10
13 23
24
24-28 30-32
35 45
50
the sharing of code, and how different program models allowed differential kinds of protection from sharing. 55--goes into Gates' hobbyist letter. 59 The culture at MIT merged with the microcomputer revolution of the 1960s and 70s and the "hobbyist" countercultural movements in the 70s and 80s (see Little Brother), and together they formed a hacker ethic that was "centered on a deeply-rooted set of beliefs about what it means to program, what it means to be an artist, even about what it means to be alive." [then he makes a generalization about hackers I don't agree with.] He draws out the tensions between the free software and open source movements. OSS (raymond) focuses on a viable business model and FS (stallman) is more political, and anti-corporate. 65-->but they do have in common "a deep and abiding sense of programming as an art, and of code as an essentially literary aesthetic form." "cybertheory appropriates the computer to serve political agendas that are in many ways antithetical to the issues with which computing culture is itself concerned; this political dislocation is itself perpetuates the divide between the two fields [of the humanities and computing]." Cybertheory sucks because it fantasizes about the computer and doesn't pay attention to its history, who's behind it, or how it works. "a nuanced, historically-inflected understanding of computing must acknowledge both the abstractionsand the aesthetic, cultural, and political effects of those abstractions." He has a "manifesto for informed theory". Compares Haraway's Cyborg Manifesto with the GNU Manifesto, in which "Stallman is building a system; Haraway is trying to dismantle one." Haraway's prose is almost impenetrable, and she has these ideas about faceless things controlling. Working from Haraway's idea that people need to recode themselves, "The ability of 'coding' to convey the idea of a cyborg playfully and effectively undermining the system from within depends on Haraway's extraordinarily unproblematic notion of code as a kind of writing that is presently in the hands of the too-entitled, too thoughtless few, the scientists, technologists, capitalists, and warmongers whose special access to code gives them corresponding access to political and economic power. " She doesn't pay attention to this movement actually happening all around her [although Black does not acknowledge the gender disparities in OSS]. 87-->"by assuming so blithely that code belongs to the victors, Haraway manages to reinforce the very equations she claims she wants to dismantle." 88-->"It secures its vision of transgressive women at the expense of a technological reality it knows nothing about, and so ironically plays into the hands of controlling corporate interests by granting it a priori all the control is desires eventually to have"
64
66
76
79
86
Chap 2 96 Programming can be conceived as literary, and has been in its history. In particular, it became considered moreso by the 1960s, when higher-level programming languages were in use. Ada Lovelace's "figural" contributions were critical to programming history (and to Black's arg). "By the time high-level programming languages came into widespread use during the 1960s, it had become customary to conceptualize programming as a form of literary composition." Programmers didn't just analogize their work as art; it was art. Ensmenger here, too--the bohemian angle on corporate structure with programmers. Black ties the push toward more managed code to the IBM System/360 programming effort, which separated the programmers from the operators of computers. [I don't know that I agree that the abstraction from hardware made the managerial push more prominent.] Now that the languages were more abstracted from the specific machine, the literary connotations began. Jean Sammet, a historian of programming languages, discusses the reasons behind the many languages written since the 1950s--it's for personal interest and style rather than any technical need. Although language designers often claim superiority over previous languages, "in the last analysis it almost always boils down to a question of personal style and taste." Knuth's behemoth and unfinished "The Art of Computer Programming" has been hugely influential in shaping the artistic framing of programming. Black asserts that "the debate about whether code is literature grounds one of the most economically significant legal struggles of our time." IP and Code. The identity / ontological issues with programming -- "Is it a machine, a process, a creative expression, or a hitherto unimagined amalgam of all three?" He quotes a 1941 decision by Justice Story about how IP acts as the "metaphysics of the law." Knuth thinks of a good program as one both aesthetically beautiful and that runs well. Black thinks its problematic that literary studies have ignored the aesthetics and texts of computer software "Simply put: by leaving the question of code's ontological status to lawyers and corporations, literary critics have thus far absented themselves from what may well be the most politically and economically profound literary debate of our times." Review of legal status of software, from 1969 unbundling from IBM, to hobbyists in 1970s,
98
100
102
103
104
106
107
109 110
and it became IP then. 1960s and 1970s, trade secret laws were used. 1972 Gottschalk v. Benson--can't patent mathematical algorithms. 111 Gov felt pressure to protect software in the 1970s. Congress convened CONTU (National Commission on New Technological Uses of Copyrighted works) in 1974. Published report in 1978. CONTU recommended that Congress define computer programs as "literary works" and protect them under copyright. Black focuses on Hersey's (dissenting) rec to allow protection of object code, but not source code. 1980 Amendment to law that computer programs can be defined as "literary works" This was great, except that code could then be reverse-engineered, and a rival company could come out with something that does the same stuff. Diamond v. Diehr showed that industrial process (using software) could be patented. 115: Flood of patents after that (Lessig) Lessig notes that programmers are almost uniformly against software patents. (Future of Ideas 208) Discussion of ownership vs. authorship in OSS projects. Closely related, but not the same. He says that Stallman "oppose[s] copyright as a matter of principle"--but that's not really true (see recent /. Article on stallman, and the reliance on copyright law) 118 Black finds it ironic that even though lit crit will read anything as a text now--bodies, clothes, garbage--they still don't see software as a text, even though it's been defined as a "literary form" for two decades. "An even greater irony is that when the almost uniformly left-wing literary critical culture neglects to consider software as a textual entity shaped by real material conditions and situation in actual political contexts, it becomes complicit with a corporate agenda that would deprive software of textual status in order to secure prohibitive patents." On Kittler: Kitt gets the abstraction principles of code, but "lacks any cultural or aesthetic understanding of programming practice.Within programming culture itself, the movement has been quite the opposite to the one Kittler emphasizes: programmers understand programming not as deconstruction of language but as a contribution to literary history." In 1970s, source code was put on CS reading lists, "it was then that computer scientists began to think of code as something whose form and content are best revealed and appreciated by way of detailed line-by-line exegesis." He claims that" close reading code" contributes to its sense of artistry, etc. (I'm not sure that's exactly connected to the literary angle, although I do think it's connected to the LITERACY angle) He ties the concept of
113
113-4
115-6
117
118
119
the "art of code" to Unix. Unix was written in C rather than Assembly, and it was shorter (<10000 lines), in 1969. 120 Unix was open and flexible (as Thompson and Ritchie write in 1974, "The Unix TimeSharing System" unlike IBM's OS/360. People liked Unix because it could be ported to different architectures, and when Thompson and Ritchie mailed out copies, it came with source code that people could and did peruse. Thompson in 1975 in Berkeley had had a Unix code-reading group. (Black calls this the beginning of "the literary culture of reading code.") Black looks at John Lions at University of New South Wales and the way he had people read code, as a literary approach, esp when he published "Commentary on the Unix Operating System". He quotes programmers talking about Unix as though it's a piece of literary history, beautifully written in C. Western Electric wanted to protect its investment (?) and Unix v. 7 released in 1979 was closed-source. In mid-80s Stallman wanted to recreate the collaborative and open Unix culture, and began to write GNU. Andrew Tanenbaum, a prof in Amsterdam, wrote a miniature Unix (Minix) and distributed it with a textbooks Operating Systems: Design and Implementation. Torvalds studied this book at university and used Minix as a starting point for his own Linux. Knuth develops the analogy of computer programs as "works of literature" -- Literate Programming, 99. Programming & code should be about the human reader. He annotated his own programs, TEX and METAFONT. Knuth wants good typesetting for code, just as for lit. His WEB system generates the commentary and machine-executable code in nice typesetting. Landow gets it wrong: art and computing wasn't a phenom of the 90s; Oulipo was part of that. (Algol) Perl poetry, started with a Haiku written by Larry Wall in 1990. Sharon Hopkins first thought of this writes perl poetry, and wrote a major essay on it. (142) No lit crit has written about Perl poetry--its criticism is provided by its practictioners (Black thinks that lit crit should address this) He contextualizes this with "codework", diff from cybertexts (acc to Mark Poster) that tend to be enacted with invisible code underlying them. Codework (defined by series of essays in American Book Review) where the technology is exposed in the body of the text. But
121
123
126
128
129
131
139
141
146
147
Black notes that the codework described in these essays are hypertext or pseudocode. 148 "In Perl poetry, the code does not underpin the poetry: the poetry is the code, and the code is the poetry." James Joyce as a "programatically aesthetic author" Summary of chapter on Joyce: Black uses the idea of OO programming to look at Joyce. Joyce had an obsession with engineering, and in FWake, Joyce had a story generating composition system that resembled OO objects. (sigra) He figured someone else could complete his novel with the help of the sigra. Also, Joyce had an interest in how children read, as did Piaget, Papert and Kay. They all (roughly) thought of kids as being able to learn better not in formal education (well for Papert, specifically constructivist ways). Interesting new method of Joyce criticism. Montfort buys the analogy, but says that it's not completely grounded: what does the text of FW compare to? The source or object code? Conc--he visits Heideigger's Question of Technology and talks about programmers merging the techne and poeisis. Claims this is the first examination of code as art.
148