data science my projects research

Husband and wife: analysing gender issues through literary big data

Some time ago a friend made me realise the peculiar distribution of the word gay in English literature: relatively common in the 1800s, then in decline, then in massive recovery after the 1970s. Of course, the word here is used with two different meanings, the first one (“light-hearted, care free”) more common in the past few centuries, with the second (“homosexual”) going mainstream in the latter part of the 20th century. All of this can be easily visualised using Google Ngrams.

I became rather curious about this because I realised that gender issues have often been written about in literature; also, the ways in which familiar scenes have been depicted could easily be a proxy to understanding the relationship between the genders, especially in their strict, unchanging view often purported by traditionalists in our society.

So I charted four words: manwoman, husbandwife. The result is enlightening.

You see, it’s not just that “man” dominates. This can be explained in many ways, especially by the common use of “man” as a synonym of “human being”. The sudden growth in the latter part of 1700 is pointing to several phenomena happening in those years, from Enlightenment to the French Revolution.

Some data points:

  • “husband” is rarely used, compared with “man”; the ratio is about 1 to 10
  • conversely, “woman” and “wife” follow a similar trend with a much smaller ratio
  • “wife” has been used more than “woman” until the late 1800
  • “woman” becomes increasingly more important than “wife” after the 1970s.

Isn’t that a rather accurate description of what happens not just in the English literary corpus but, more widely, in society?

policy politics research

Research and democracy

This is the content of my letter published by the New Scientist.

It refers to Dan Hind’s proposal, on a previous issue, to make research topics subject to public scrutiny in order to create a “democracy of research” free from the action of lobbies. I suspect this is dangerous at worst and naive at best, as this would make the lobbies’ work much easier.

Hind was not available to comment.

From Giuseppe Sollazzo, London, UK

As much as it is true that public scrutiny is the base of every democratic system, I’m not sure that this concept can be easily applied to research anywhere but in an ideal world. Exclusion of people from the voting system based on their level of education would be considered anti-democratic, but what happens when the electorate is ignorant?

In the US, the incoming Republican House majority leader, Eric Cantor, has instigated a public vote that has already favoured cuts in science funding over other areas and is now being used to determine where these cuts should be made (11 December 2010, p 7).

Would Hind let people who voted for Cantor, Sarah Palin and the like decide how to allocate research funds? If not, the democracy is flawed. If he does, good luck to the rest of us.