The jury is still out on whether homosexuality is determined at birth. However, a recent study by researchers at Stanford University suggests it is written on the faces of gays and lesbians.

In an article published earlier this year in the Journal of Personality and Social Psychology Yilun Wang and Michal Kosinski used big-data techniques to analyse 300,000 images from dating sites of 75,000 men and women, both straight and gay. On average, they found, gay men have more feminine faces than straight counterparts and lesbians display more masculine features than heterosexual women.

Using a deep neural network, they claimed that they could correctly distinguish between homosexual and heterosexual men 81 percent of the time from a single image, and 71 percent of the time for women. The results were even more accurate if they analysed five images.

As Ulrich Kutschera, of the University of Kassel, in Germany, points out in a recently published article in the Japan Journal of Medicine, this is good news for the ancient art of physiognomy. As long ago as Aristotle, in Ancient Greece, physiognomy was believed to be scientific, because “our character and mental capabilities are completely written down in our faces”.

The philosopher’s views were given new life by a writer of the Italian Renaissance, Giambattista della Porta, who published a book titled De Humana Physiognomonia. In the 18th century a Swiss poet and pastor, Johann Casper Lavater, wrote a massive four-volume work on interpreting faces which became very popular in Europe, partly because of its handsome illustrations. And in the 19th Century, the Italian criminologist Cesare Lombroso claimed that he was able to detect criminal tendencies by measuring the shape of heads.

Even the great biologist Charles Darwin thought that there might be something in physiognomy – not in its vulgar version, á la della Porta and Lavater, but in the fact that animals and humans sometimes express emotions in similar ways. It supported his views on evolution.

He had reason to distrust the popular version – he almost did not sail on the Beagle because the ship’s captain, one Robert FitzRoy, an ardent enthusiast of physiognomy, believed that “no man with such a nose could have energy”. Fortunately, the shape of Darwin’s brow saved him. He went on to a life of feverish activity and accomplishment, while FitzRoy, sadly, committed suicide in his retirement.  

Nowadays, physiognomy is regarded as a bogus science, a superstition which flourishes only on day-time TV shows.

However, as Kutschera points out, the Stanford research suggests that “a ‘a kernel of truth’ is hidden behind this old philosophical idea”.

Is this good news for gays and lesbians? Media coverage of the article suggested that it was because it supports the notion that homosexuality is an innate orientation over which they have little control.

But there are dangers. Governments and law enforcement officials everywhere are adopting surveillance techniques based on big data. If Wang and Kosinski’s claims are confirmed, it would be easy for them to identify gays and lesbians by their facial features. In most Western countries, this might not be a problem.

But elsewhere? In Iran, Mauritania, Saudi Arabia, and Yemen the punishment for homosexual activity is a death sentence. China is already exporting its highly sophisticated surveillance technology to Zimbabwe where homosexual activity is illegal and despised. The consequences for gays and lesbians could be disastrous. As Wang and Kosinski point out, “the predictability of sexual orientation could have serious and even life-threatening implications to gay men and women and the society as a whole.”