Hi John and Bernhard (and all)

Good point on the [PaSS] mail subject thing - will see if I can do it automatically through JISC, but if not, it's easy enough to type out and very helpful in terms of organising an inbox!

While I'm very much on board with Galloway's strategy of having a critical stance on different usages and applications of digital data and algorithms (especially in terms of drawing out political aspects), this thread brought to mind Mair, Greiffenhagen and Sharrock's paper on how statisticians do their work http://journals.sagepub.com/doi/abs/10.1177/0263276414559058 - this paper looks at the practices of "pure statisticians" who were tasked with producing representations of data for a social science project, noting that the statisticians needed to a good deal of "qualitative" work to understand the social significance of the data they were given and instructed to represent, as well as to understand the recipients of their statistical outputs (in terms of what those people needed to be able to see in those outputs). This is not to say that the statistics work they did was bad stats or that they were trying to fiddle the numbers or anything, but just that the things we often assume are "purely quantitative" are, in the real world of practice, not done that way. And this is of course a good thing - how else would the statisticians produce statistical information that effectively connects with social science questions? So I guess it would be nice to have this pointed out in the context of Galloway's conflation of digital research with positivistic and quantitative methods - one reading of Mair et al's paper would be that Galloway's definition probably doesn't hold when you actually look at the work "new algorithmic industries" and the people who work for them actually do, on a day-to-day "in the lab" type basis, to generate and make sense of their data. Naturally, I'm not trying absolve Amazon or any other corporation of their responsibilities in terms of data collection and handling :-) but just thinking that it would be interesting to SEE (rather than assume, as Galloway does) the reasons and decision-making and practices that go into that work.

Cheers

Phil



________________________________________
From: Programming as Social Science (PaSS) <[log in to unmask]> on behalf of Bernhard Rieder <[log in to unmask]>
Sent: 17 March 2017 12:04
To: [log in to unmask]
Subject: Re: cybernetic hypothesis

Hi,

First, thanks for creating this list! One small suggestion: adding [PaSS] to the mail subject would make email handling *a lot* easier, for me at least.

Second, a quick introduction, my name is Bernhard, I’m not really a social scientist, but I write software (http://labs.polsys.net) that is used by a lot of social scientist. My own research interest is mainly software itself, especially stuff like information retrieval and machine learning, which I look at mainly from a humanities/economics perspective.

Concerning John’s points and the Galloway quote, I’m reacting to this since I find that it’s a really important point. I do agree with Galloway’s “state-of-affairs” assessment, but I can’t help feeling that, in a sense, software has been “abandoned” to the positivistic and the market forces he decries. The way I see it, computation in itself has nothing positivistic too it, much like mathematics can be seen as pure axiomatic formalism, as games we play in symbolic systems. In graph theory, something like a centrality measure is a mere formal transformation of some data. The problems begin when we bind the results back into some social reality and consider that the calculation allows us to say something about that social reality, e.g. something about status, authority, influence, etc. This is, I think, where the positivism happens. But there is so much space for doubt, prudence, and critique at that interface.

What I see, however, is that those “critical voices” (like Galloway) who would favor epistemological alternatives to positivism miss that space of interpretation, that margin where one could use the calculative capacities of the computer in ways that do not mean subscribing to a mindset where reality is reducible to formal representation. The critical voices have too little technical knowledge to imagine different uses of computing, the result being the classic binary reaction of total rejection or acceptance (interestingly, some of the scholars I work with are perfect positivists although they have no quantitative/computational understanding at all). That’s what I mean by “abandoned” - there are too few people that champion a different view of computation as an intellectual tool.

But critical practice could start at a really basic level by questioning methodological or number fetishism when it’s appropriate. A stupid example: I have stopped using fractions in talks because they suggest a level of exactitude that few methodological setups can actually deliver. Using grey fonts for numbers is another thing I have tried out as a conversation starter. This is of course very basic, but I think it’s necessary to highlight the methodology on all levels, to make it a subject and to discuss it in nuanced terms, where we do not fall into the binary reaction I mentioned further up.

TL;DR: we need more computationally competent people to show that computation, as an intellectual tool, does not have to follow positivist epistemologies

Sorry for the rambling,
Bernhard


> On 17 Mar 2017, at 11:21 , Boy, John <[log in to unmask]> wrote:
>
> Because I'd hate to see conversation on this list simple peter out, I thought I'd send around a little a quote from a piece by media theorist Alex Galloway in hopes of provoking further exchange. I recommend reading the whole piece [1], which mostly addresses digital humanities but I think also has important implications for the idea of "programming as social science".
>
> He writes: "But beyond the challenge of unequal talent and resources [among humanists as compared to corporations like Amazon] is the question of critical efficacy. Is it appropriate to deploy positivistic techniques against the self-same positivistic techniques? In a former time, such criticism would not have been valid or even necessary. Marx was writing against a system that laid no specific claims to the apparatus of knowledge production itself -- even if it was fueled by a persistent and pernicious form of ideological misrecognition. Yet, today the state of affairs is entirely reversed. The new spirit of capitalism is found in brainwork, self-measurement and self-fashioning, perpetual critique and innovation, data creation and extraction. In short, doing capitalist work and doing intellectual work -- of any variety, bourgeois or progressive -- are more aligned today than they have ever been. Hence there appears something of a moral crisis concerning the very validity of scholarly methodologies. Such methods are at best underfunded and impotent cousins to the new algorithmic industries and at worst unknowing shills for that same system of canalization and debasement. The question is no longer 'can we use the master's tools to take down the master's house?' The question is 'can we still use our own tools now that the master has taken them up?'" (p. 110)
>
> Perhaps the "critical efficacy" of our intellectual pursuits is not equally important to everyone on this list, though I think he's right that we at least need to consider our role in current moral crises.
>
> But I think Galloway raises a few interesting questions even beyond that. In particular, I wonder whether he correct to conflate digital research techniques with "positivistic" and "quantitative"?
>
> I look forward to reading your thoughts!
>
> John
>
>
> PS: As I was writing, a message from Phil came through the list. Yay!
> PPS: I'm still hanging out in IRC channel irc.oftc.net/#pass in case anyone wants to join :)
>
> 1: Galloway, Alexander R. 2014. The cybernetic hypothesis. Differences: A Journal of Feminist Cultural Studies, vol. 25, no. 1, pp. 107-131. https://doi.org/10.1215/10407391-2420021