Dreaming in Code
Dreaming in Code
DREAMING IN CODE
T
his morning, floating through that state between sleep
and consciousness in which you can become aware of your
dreams as dreams immediately before waking, I realized that
I was dreaming in code again.1 This has been occurring on
and off for the past few weeks; in fact, most times I have become aware
of the content of my unconscious mind’s meanderings, it has been
something abstractly connected with my job. I remember hearing the
sound of the call centre in my ears as I drifted in and out of sleep when I
was working there, and have heard stories from friends of doing an extra
shift between going to sleep and waking—the repetitive beeps of a super-
market checkout punctuating the night. But dreaming about your job
is one thing; dreaming inside the logic of your work is quite another. Of
course it is unfortunate if one’s unconscious mind can find nothing bet-
ter to do than return to mundane tasks, or if one’s senses seem stamped
with the lingering impression of a day’s work. But in the kind of dream
that I have been having the very movement of my mind is transformed:
it has become that of my job. It is as if the repetitive thought patterns and
the particular logic I employ when going about my work are becoming
hardwired; are becoming the default logic that I use to think with. This
is somewhat unnerving.
1
This essay was originally published in Swedish as ‘Att arbeta i sömnen’ in Dissident,
no. 3, 2008. A longer version will appear in Endnotes 2.
lucas: Work 127
manager’, and I primarily deal with those on my side of the great divide,
it is even possible to develop a certain ‘us against them’ attitude.
From our point of view, business and its needs appear as parasitic exter-
nalities imposed upon the real functioning of our use-value-producing
enterprise. We are strangely tied to a certain normativity; not just that of
doing the job right in a technical sense, but also that of thinking in terms
of the provision of real services, of user experiences, and of encouraging
the free flow of information. This sometimes spills over into outright
conflict: when business advocates some tortuous use of language to
hype ‘the product’, the techies will try to bend the stick back towards
honesty and transparency. ‘What goes around comes around’ seems
to be the prevalent attitude in web development in the era after ‘Web
2.0’: provide the services cheap or free, give away the information, be
decent and hope that somehow the money will flow in. If business acts
with the mind of money capital, encountering the world as a friction or
recalcitrance which it longs to overcome, and if a tendency to try to sell
snake oil follows from that, in the strange world where technical pride
opposes itself to capital as capital’s own developed super-ego, use-value
rules with a pristine conscience; everything is ‘sanity checked’—to use
the terminology of my boss—and the aggregation of value appears as an
accidental aside.
There is little space left in this relation for a wilful ‘refusal of work’:
given the individually allocated and project-centred character of the job,
absenteeism only amounts to self-punishment, as work that is not done
now will have to be done later, under greater stress. Apart from that,
128 nlr 62
there is the heavy interpersonal pressure that comes with the role: since
a majority of the work is ‘collaborative’ in a loose sense, heel-dragging or
absenteeism necessarily involves a sense of guilt towards the technical
workers in general. Nor is sabotage a creative option here; not because
of the supposed pride of the skilled worker, but due to the nature of the
product. On a production line, sabotage may be a rational tactic, halt-
ing the relentless flow to provide half-an-hour of collective sociability.
When one’s work resembles that of the artisan, to sabotage would be to
make life harder. Occasionally one hears of freelancers or contractors
who write confusing and idiosyncratic ‘spaghetti code’ in order to keep
themselves in work. This technique may make sense when a company
relies heavily on particular individuals; but in a typical development
team, which uses feedback-centred it management methodologies such
as ‘agile’ and ‘extreme’ programming, and where ‘ownership’ of a project
is always collective, high-quality, clearly readable code has a normative
priority that goes beyond whatever feelings one might have about doing
one’s job well.
II
Just as capital posits its own constraint in the form of the state, in order
not to destroy itself through the rapacious self-interest of each individual
capital, so—after an early period of ugly coding, due to the fragmenta-
tion of the internet into a babel of different platforms, browsers and
languages—a consensus formed in the world of web development that
‘standards’ were important. Central to this was an idea of universalism:
anything that adheres to these standards should be supported; anything
that does not is asking for trouble. Microsoft became a pariah among
web professionals due to its continuing contempt for standards and its
penchant for developing proprietary annexes on the public space of the
net. Developers proudly sported web-standards badges on their personal
sites and became vocal advocates of technologies like Mozilla’s Firefox
which, as well as being ‘open source’, beat Internet Explorer hands down
in terms of standards-compliance; in fact, in the moral universe of the
130 nlr 62
III
I have enough difficulty dealing with the stuff when I am awake, and I
suspect my unconscious mind fares little better. But sometimes it does
have a meaning. One morning recently I was woken by the thought of
a bug in some code that I had written, which I had not realized was
there. My sleeping mind had been examining a week’s work and had
stumbled upon an inconsistency. Since I am a thought-worker, and since
the identification and solution of such problems is a major aspect of my
job, it is not fantastical to say that I had been performing actual labour
in my sleep. This is not the magical fecundity of generalized creative
power, churning out ‘value’ beyond, and ontologically before, the labour
process. It is actual work for capital, indistinguishable in character from
what I do during my working day—but occurring in my sleeping mind.
Suddenly the nightmarish idea of some new kind of subsumption—one
that involves a transformation of the very structures of consciousness—
looms into view. Indeed, I find that standard patterns of thought seem
increasingly burned into my mind: the momentary recognition that
there is a problem with something prompts a fleeting consideration of
the possible locations of faulty code, before I consciously jolt my mind
out of code-world and into the recognition that ‘bugfixing’ cannot solve
all problems. There is something terrifying here.