I gave a talk at Google and witnessed a perfect example of how badly broken the culture of the tech industry is today – Business Insider


Google UK office
The view from Google in London, where the author gave a
talk.

Google
Earth


Five years ago, Silicon Valley was rocked by a wave of “brogrammer” bad behavior, when overfunded,
highly entitled, mostly white and male startup founders did
things that were juvenile, out of line and just plain stupid.

Most of these activities – such as putting pornography into
PowerPoint slides – revolved around
the explicit or implied devaluation and
harassment
 of women and the assumption that
heterosexual men’s privilege could or should define the
workplace.

The recent “memo” scandal out of Google shows how far we
have yet to go.

It may be that more established and successful companies don’t
make job applicants deal with “bikini shots” and “gangbang
interviews.”
 But even the tech giants
foster an environment where heteronormativity and male privilege
is so rampant that an engineer could feel
comfortable writing and distributing a
screed
 that effectively harassed all of his
women co-workers en masse.

This is a pity, because tech companies say they want to change
this culture. This summer, I gave a talk at Google UK about my
work as a historian of technology and gender. I thought
my talk might help change people’s minds about women in
computing, and might even help women
and nonbinary folks working at
Google now.

Still, the irony was strong: I was visiting a multibillion-dollar
tech company to talk about how women are undervalued in tech, for
free.

Going to Google UK

I went to Google UK with significant trepidation. I was going to
talk about the subject of my upcoming book, “Programmed Inequality,” about
how women got pushed out of
computing
 in the U.K. In the 1940s through
the early 1960s, most British computer workers were women, but over the
course of the ’60’s and ’70’s their numbers dropped as women were
subjected to intentional structural
discrimination
 designed to push them out of
the field. That didn’t just hurt the women, either –
it torpedoed the once-promising British computing industry.

In the worst-case scenario, I imagined my talk would end with a
question-and-answer period in which I would be asked to face
exactly the points the Google manifesto made. It’s happened
before – and not just to me – so I
have years of practice dealing with harsh critics and tough
audiences, both in the classroom and outside of it.

As a result of that experience, I know how to handle situations
like that. But it’s more than just disheartening to have my work
misunderstood. I have felt firsthand the damage the phenomenon
called “stereotype threat” can wreak on women: Being
assumed to be inferior can make a person not only feel inferior,
but actually subconsciously do
things
 that confirm their own supposed
lesser worth.

For instance, women students do
measurably worse on math exams
 after
reading articles that suggest women are ill-suited to study math.
(A related phenomenon, impostor syndrome, runs rampant through academia.)

A surprising question from a female Google engineer

As it happened, the audience was familiar with, and interested
in, my work. I was impressed and delighted with the caliber and
thoughtfulness of the questions I got. But one question stood
out. It seemed like the perfect example of how the culture of the
tech industry is so badly broken today that it destroys or
significantly hinders much of its talent pool, inflicting
stereotype threat on them in large numbers.

A Google engineer asked if I thought that women’s biological
differences made them innately less likely to be good engineers.

I replied in the negative, firmly stating that this kind of
pseudoscientific evolutional psychology has
been proven incorrect at every turn by history,
and that biological determinism was a dangerous cudgel that had
been used to deprive black people, women and many others of their civil
rights – and even their lives – for centuries.

The engineer posing this question was a woman. She said she felt
she was unusual because she thought she had
less emotional intelligence and
more intellectual intelligence than most other women, and those
abilities let her do her job better. She wondered if most women
were doomed to fail. She spoke with the uncertainty of someone
who has been told repeatedly that
“normal” women aren’t supposed to do what she does, or be who she
is.

I tried to empathize with her, and to make my answer firm but not
dismissive. This is how structural discrimination works:
It seeps into all of us, and we are barely conscious of it. If we
do not constantly guard ourselves against its insidious effects –
if we do not have the tools to do so, the courage to speak out,
and the ability to understand when it is explained to us – it can
turn us into ever worse versions of ourselves.

We can become the versions that the negative stereotypes expect.
But the bigger problem is that it doesn’t end at the level of the
individual.


women in tech
Marie Hicks

A problem of structure

These misapprehensions bleed into every aspect of our
institutions, which then in turn nurture and (often unwittingly)
propagate them further. That was what happened
when the Google manifesto emerged, and in the media
frenzy that followed.

That the manifesto was taken as a potentially interesting or
illustrative opinion says something not just about Silicon
Valley, but about the political moment in which we find
ourselves. The media is complicit too: Some media treated it as
noteworthy only for its shock value.

And others, rather than identifying the screed as an example of
the writer’s misogyny, lack of historical understanding, and
indeed – as some computer professionals have pointed
out
 – lack of understanding of the field of
engineering, handled the document as a think piece deserving consideration and
discussion
.

The many people who said openly
and loudly that it was nothing of the sort are to
be commended. But the fact that they had to waste time even
addressing it shows how much damage casual, unreflective sexism and
misogyny
 do to every aspect of our society
and our economy.

The corporate response

Google, for its part, has now fired the writer, an expected move after
the bad publicity he has helped
rain down on the company. But Google has also – and in the very
same week that I gave my talk there – refused to comply with a
U.S. Department of Justice order to provide statistics on how it paid its women
workers
 in comparison to men. The company
claims that it might cost an estimated US$100,000 to compile that
data, and complains that it’s too high
a cost for their multibillion dollar corporation to bear.

The company will not expend a pittance – especially in relation
to its earnings – to work to
correct allegedly egregious gender-biased salary disparities. Is
it any surprise that some of its employees – both men and women –
view women’s contributions, and their very identities, as
being somehow less inherently
valuable
 or well suited to tech? Or
that many more silently believe it, almost in spite
of themselves?

People take cues from our institutions. Our governments,
corporations, universities and news
media shape our
understandings and expectations of
ourselves
 in ways we can only partially
understand without intense and sustained self-reflection. For the
U.K. in the 20th century, that collective, institutional
self-awareness came far too late to save its tech sector.

Let’s hope the U.S. in the 21st century learns something from
that history. At a time when technology and governance are
increasingly converging to define who we are as a nation, we are
living through a perfect – if terrifying – teachable moment.

Marie Hicks is an assistant professor of history at the
University of Wisconsin-Madison.

I gave a talk at Google and witnessed a perfect example of how badly broken the culture of the tech industry is today – Business Insider

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>