Because I’m human, and susceptible to the weaknesses and vanity of our species, I google myself with some regularity. Not every week, but maybe, you know, twice a year. Maybe once a quarter. Look, I will not be shamed for this, it’s a reasonable thing to do, like going to the dermatologist to check your moles.
For a few years, I’ve noticed something odd in the suggested search options that display either as a dropdown in the search box or at the bottom of the page under the “searches related to” heading.
According to Google, the suggested searches are based on what real people actually typed into the search box. They are not influenced by the availability of the actual results — meaning that it doesn’t matter that there are very few results for “katie notopoulos wedding,” none of which are relevant.
The company says auto-suggestions are based solely on what other humans actually searched for. Well, almost. There’s also some personalization based on location, so if you search “best pizza,” you’ll see a suggestion for “best pizza in [your town].” They’re also influenced by your previous searches and what you’ve clicked on in the past.
Safiya Umoja Noble, associate professor of information studies at UCLA who has written about race and sex bias in Google results, is somewhat skeptical. “It wants you to click on these things. These suggestions are going to lead you to more optimal content for Google,” she told BuzzFeed News. In her book Algorithms of Oppression, she described how Google responded to a series of embarrassing incidents when horrible racist suggestions kept popping up around certain search terms. Google has largely dealt with these cases by manually removing them.
But these marriage-related suggestions aren’t blatantly and offensively sexist — they’re just, well, mildly creepy and passively sexist. Not worth a takedown from Google. And while the suggestions are free of SEO-gaming algorithms, they are still the result of some form of machine learning, which is susceptible to bias. “It’s untenable to think that what shows up in search engines is the truth about someone,” said Noble.
Auto-suggestions become self-fulfilling prophecies.
It’s possible that these auto-suggestions become self-fulfilling prophecies. Google confirmed to BuzzFeed News that if someone clicks on one of the suggested search phrases, it registers as a new search. With people like the groups I picked, there isn’t much overall search volume — we aren’t celebrities. With just a few clicks from people who see “wedding” as a suggestion, and Google thinks this is an important and relevant term, and it stays as a suggestion.
I attempted to manipulate this. I asked people on Twitter to click on a link to the search results for “katie notopoulos pulitzer” to see if a sudden uptick in searches would become the top suggested search when you type in my name. To my dismay, it didn’t work — nor have I won the Pulitzer (yet). My best guess is that it got flagged as inauthentic behavior (Google told me they could not confirm how they detect manipulation to prevent bad actors from beating their system).
I’m not a scientist. I’m just a tech blogger, standing in front of some Google results, asking why this reeks of patriarchy.
First, I needed to find out if what I was anecdotally seeing really was true. To test my theory, I tried three cohorts: married BuzzFeed News staff, Harvard history professors, and the 2019 Forbes’ 30 Under 30 list in the health care category. I chose these groups because academics and 30 Under 30 winners are similar to journalists in the sense that they’re not actually famous, but they do have a lot of Google results for their names. These are also semipublic figures, known for their professional work rather than personal lives.
To avoid my own search preferences tainting the results, I used a new browser that I don’t use at all — Firefox — in a private search tab, not logged into Google. I already knew the marital status and gender of my colleagues, but for the Harvard professors and 30 Under 30 list, I assigned gender based on first names and photos on the website. That’s an imperfect method, but it was sufficient for this relatively small sample.
For married women editors and reporters who work at BuzzFeed News, 6 out of 16 had some variation of “husband” or “wedding” as a suggested search term. Only 2 of 20 married men had “wife” as a suggested search.
The Harvard history department has 52 professors listed on its website (not including adjuncts and visiting staff) — 29 men and 23 women. Four of 29 male professors had a suggestion of “spouse,” whereas 11 of 23 female professors did. That’s 13% of male professors and 47% of female professors.
One particularly interesting example was a married couple in the department. When searching the woman’s name, her husband’s full name was suggested, but the male professor had no marriage-related suggestions.
Last, I tried the winners of the Forbes 2019 30 Under 30 list for the health care category (to avoid actual celebs). Only 1 of the 25 men had a marriage-related suggestion; 4 out of 15 women did.
Of course, my experiment is not scientific or rigorous — it’s back of the envelope. There are a lot of factors not accounted for, like how many people are actually married, and the whole experiment implies cisgender heterosexuality — not to mention its small sample size. A real social scientist would never let this pass muster. But I’m not a scientist. I’m just a tech blogger, standing in front of some Google results, asking why this reeks of patriarchy.
So what are we to make of this? It’s a sociological question, not a technological one: Why would someone google a public-but-not-famous person’s spouse in the first place?
My initial thought is that it’s fueled by horniness — thirsty creeps trying to find out if women are single or married. But as much as it flatters my ego to imagine this, my lived experience informs me that this is not the case. I am on the wrong side of 30 for strangers on the internet to get horny for (some of my younger colleagues told me that they do get horny DMs from readers; this is certainly a real thing that happens). And thirst for professors is of course a real thing; the chili pepper rating for hotness on RateMyProfessor.com is a testament to that unsavory aspect of higher education.
But if horniness was the main motivator here, you’d expect to see other similar horny suggestions for unmarried women — “jane smith single” or “jane smith dating” — but you don’t. The search behavior here isn’t that of someone who wants to find out if someone is married or not — that search would probably look more like “jane smith married.” The search behavior is more congruent with someone who knows the person is indeed married, and wants to know who that spouse is.
It’s normal and non-horny to be curious about who people are married to. That’s why it's interesting to find out who your high school classmates are married to, or meet your coworkers’ partners at the company holiday party. Being curious about strangers’ weddings is also normal — that’s why there’s a whole section of the newspaper dedicated to it each Sunday. But it doesn't explain the gender disparity.
Perhaps the most likely reason for why someone searches “jane smith husband”: The husband is also a notable person, maybe even in the same industry. A few of the people at BuzzFeed News where “husband” or “wedding” popped up as a suggestion are married to other journalists.
However, this only goes one way with gender. When I searched the names of those journalist husbands of BuzzFeed News reporters, “wife” wasn’t a suggestion. A few male reporters here are married to women journalists, and while there were no marriage-related suggestions for the male reporter, when I searched their wives’ names, “husband” or the husband’s full name was a suggestion.
For Harvard professors, it was often the case that the husband was also a professor or otherwise notable in his professional field, like a federal judge. It’s not a shock to me that women at the top of their fields — tenured Harvard professors — might be married to other high-achieving professionals.
But it suggests to me a clue about why the gender reversal doesn’t work. High-achieving professional men may have gotten to their position by having a spouse who put aside or deprioritized her career.
Ultimately, without evidence that points to one cause, I am making my best guess here: Our culture prioritizes women’s private and family lives over or equal to their professional lives in a way that it simply does not about men. We define women, even accomplished professional women, by who their husbands are. This isn’t some provocative or radical idea — this is Taylor Swift–level Feminism 101. You already knew this was going to be the answer to “why does Google keep suggesting ‘husband’ when I search for women’s names?” without having to read this whole article. I’ve basically wasted your time. Sorry.