[Humanist] 29.425 the strongest attack; reading scholarship in software

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Tue Oct 27 07:45:32 CET 2015


                 Humanist Discussion Group, Vol. 29, No. 425.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

  [1]   From:    Paul Fishwick <metaphorz at gmail.com>                      (104)
        Subject: Re:  29.422 reading scholarship in software?

  [2]   From:    "WILDER, COLIN" <WILDERCF at mailbox.sc.edu>                 (58)
        Subject: RE:  29.421 the strongest attack


--[1]------------------------------------------------------------------------
        Date: Mon, 26 Oct 2015 08:44:27 -0500
        From: Paul Fishwick <metaphorz at gmail.com>
        Subject: Re:  29.422 reading scholarship in software?
        In-Reply-To: <20151026064755.8BAF46CA9 at digitalhumanities.org>


Dear Willard:

> On Oct 26, 2015, at 1:47 AM, Humanist Discussion Group <willard.mccarty at mccarty.org.uk> wrote:
> 
>                 Humanist Discussion Group, Vol. 29, No. 422.
>            Department of Digital Humanities, King's College London
>                       www.digitalhumanities.org/humanist
>                Submit to: humanist at lists.digitalhumanities.org
> 
> 
> 
>        Date: Sat, 24 Oct 2015 10:44:46 +0100
>        From: Willard McCarty <willard.mccarty at mccarty.org.uk>
>        Subject: scholarship in software and reading the machine
> 
> 
> From time to time various of us have contemplated and a few written 
> about the problem of gaining scholarly recognition for work done in 
> software. To date, I would suppose, this is rare -- and understandably 
> so. It's not all that hard, given sufficient time, to assess the utility 
> of a digital resource for scholarship in affected disciplines, but this 
> is not the same as weighing its worth *as* scholarship. 

We need to define “scholarship.” I suggest that this word refers to
the empirical approach that we take to build upon related work. This takes
the form of bibliographies, references, and the process of developing
an argument.

Let’s consider the situation in engineering schools since that may 
provide one example for this discussion. By “engineering”, I am including
“computer science” and a host of other disciplines, since these are
often housed in schools or colleges of engineering. A tenure-track faculty 
member is expected to be a scholar, but this term is rarely used in casual
conversation. Instead, “research” is used for an identical meaning. This is
something I’ve noted with my humanist colleagues: “scholar” is part of
their vocabulary, where as in engineering, it is assumed you are a scholar
if you are research faculty, and so no need to use that term.

Consider you are a computer science faculty member. Unless you are a theorist
(e.g., theory of computing), you are expected to include software as
part of your argument. Most PhD theses include written arguments
employing significant effort in framing the arguments against related
work done in the past. The software an integral part of these
arguments, but would not be considered an example of scholarship by 
itself. However, even though writing software (sans argument) would not
be considered scholarship, significant software implementations are
considered highly significant. Their significance, as always, depends
on the tenure and promotion committee makeup. I treat
development of software and hardware as extremely important in engineering 
scholarship.

> To do this you have to be able to answer the question Mike 
> Mahoney used to ask: how do we read a machine?

By examining its design, which generally means “reading” diagrams
of increasing complexity.

> He made the question 
> more interesting and difficult by pointing out that software can only be 
> fully known in action. You know it by seeing it work, or at least you 
> cannot know it adequately otherwise.

Just like biology. Looking at cells under a microscope does not provide
information for the systemic behavior.

> A closely related question is brought up by Larry Owens' fine study, 
> "Vannevar Bush and the Differential Analyzer: The Text and Context of an 
> Early Computer", Technology and Culture 27.1 (1986): 63-95. This is how 
> he asks it:
> 
>> How does one tell the story of a machine? On what categories should
>> the analysis rest, within what interpretative framework should one
>> search for the meaning of engineering artifacts? However the
>> historian chooses to answer these questions, utility must certainly
>> play a role.... But there are other categories than utility, or,
>> maybe, broader sorts of utility than so far invoked in our
>> account.... [M]achines exist not only as tools, but also as symbols.
>> Bush's analyzers did indeed do more than simply compute x(t). To
>> flesh out our story of this particular machine, we must discover what
>> this something else was. (p. 85)
> 
> Owens goes on to open up the engineering culture of the first half of 
> the 20th Century. He looks forward to the death of interest in Bush's 
> Analyzer as digital computing took over. After what Owens calls the 
> "autopsy" of the programme that supported it, in the Spring of 1950, 
> Warren Weaver wrote to Samuel Caldwell, director of the Center where it 
> was developed,
> 
>> "[I]t seems rather a pity not to have around such a place as MIT a
>> really impressive Analogue computer; for there is vividness and
>> directness of meaning of the electrical and mechanical processes
>> involved ... which can hardly fail, I would think, to have a very
>> considerable educational value. A Digital Electronic computer is bound
>> to be a somewhat abstract affair, in which the actual computational
>> processes are fairly deeply submerged." (p. 66)

I love this quote, and included it in my paper on why computing is
an empirical science: http://dl.acm.org/citation.cfm?id=2601391

I am also a fan of analog computing, and this is a topic for a wider
discussion of how we, as historians and as a society, think about
the nature of computing.

> As Bush tirelessly pointed out "analogy machines" like the Analyzer were 
> in fact so easy to read that one could learn the calculus from them; 
> they provided a visible language expressing the "innate meaning" of the 
> mathematics that expressed the innate meaning of the processes it 
> represented. Not so for digital machines.
> 
> All of which, it seems to me, give us an inspiring example and 
> highlights just how difficult reading our digital machines is.

This is one reason why the computational thinking crowd (I count
myself as one) pushes for the importance of everyone to do some type
of coding or modeling.

-paul





--[2]------------------------------------------------------------------------
        Date: Mon, 26 Oct 2015 17:29:21 +0000
        From: "WILDER, COLIN" <WILDERCF at mailbox.sc.edu>
        Subject: RE:  29.421 the strongest attack
        In-Reply-To: <20151026064341.22E5D6CA9 at digitalhumanities.org>


Chris,

This  is a great idea. Perhaps solutions be accepted in a Turing-Test-manner - if one cannot tell the difference between a human solution and a computational solution, then it counts as solved?

Colin Wilder



-----Original Message-----
> From: humanist-bounces at lists.digitalhumanities.org [mailto:humanist-bounces at lists.digitalhumanities.org] On Behalf Of Humanist Discussion Group
> Sent: Monday, October 26, 2015 2:44 AM
> To: humanist at lists.digitalhumanities.org
> Subject: [Humanist] 29.421 the strongest attack

                 Humanist Discussion Group, Vol. 29, No. 421.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org



        Date: Sun, 25 Oct 2015 11:03:01 +0100
        From: Jan Christoph Meister <jan.c.meister at gmail.com>
        Subject: Re:  29.420 the strongest attack? - Hilbert for DH
        In-Reply-To: <20151025082555.ACA146B3A at digitalhumanities.org>


Willard,

why not scale this a bit: formulate a Hilbert-like set of seemingly unsolvalble problems? We could compile this in the good old 18th century academy fashion by issuing a challenge and offering a prize.

Best

Chris

Am 25/10/2015 um 09:25 schrieb Humanist Discussion Group:
>                   Humanist Discussion Group, Vol. 29, No. 420.
>              Department of Digital Humanities, King's College London
>                         www.digitalhumanities.org/humanist
>                  Submit to: humanist at lists.digitalhumanities.org
>
>
>
>          Date: Sun, 25 Oct 2015 08:10:37 +0000
>          From: Willard McCarty <willard.mccarty at mccarty.org.uk>
>          Subject: the strongest attack?
>
> I would be very grateful for recommendations of what members of 
> Humanist regard as the strongest, most persuasive, closely reasoned 
> attack on the use of computational methods for the analysis and/or 
> classification of literature in English. This, as I expect you know, is no straw-man.
>
> Most useful to me would be something not so focused on a single piece 
> of scholarship or single author that it might seem to have problematic 
> relevance to others. The attack would be best for my purposes if it 
> has not been adequately answered and so continues to be provocative.
>
> Many thanks.
>
> Yours,
> WM

-- 

Prof. Dr. Jan Christoph Meister
Universität Hamburg
Institut für Germanistik
Von-Melle-Park 6
20146 Hamburg
+49 40 42838 2972
+49 172 40 865 41






More information about the Humanist mailing list