See how to debunk viral photos in seconds using image search
March 3, 2014 22 Comments
Back in August I wrote about how Google Image Search and other reverse image search engines are a valuable tool to debunk viral hoaxes. Last week yet another example popped up that shows just how easy these types of debunks can be.
The new example involves social media posts about animal rights and animal testing. Photos of suffering animals are always compelling, and often go viral. While most people sympathize with the animals pictured, there is a secondary lesson here – don’t forget to apply skepticism to viral content even when the message confirms your own beliefs and pet causes.
I’m an animal fan myself – we have both a dog and a cat in our household. The purpose of this post is not to criticize animal rights activists, but show how to verify photos. So lets see how it’s done.
Warning: If you are particularly sensitive to pictures of animals in medical situations, you might not want to see the photo in this post. Try reading my previous post on this topic instead.
The most recent example was from one of those innumerable “amazing photos” Twitter accounts. These accounts have received a great deal of criticism lately for misrepresenting history and using copyrighted photos improperly. Here is the Tweet in question:
As you can see this post got an amazing number of retweets and favorites. But if you can get past your sympathy for those cats for a moment, a mere three clicks of your mouse can show you the photo is not what it seems to be.
Using Google’s Chrome browser, all you need do is right-click the photo in the tweet, and send it to Google Image Search to find out where else it has been used. Here’s what it looks like in an animation:
As you can see with this image (at the time I am writing this post) Google Image Search immediately comes up with “neuter cat” as a likely search term for this image. A scroll down and a click shows this is an image from 2011 of veterinarians at the University of Florida. Not only that, but one of the results you see scroll by is a debunk of the image at the site WafflesAtNoon, dating from last May.
A more recent debunk of this post this week at the Speaking of Research blog explains this was actually an adopt-a-thon in Gainesville, Florida in August 2011. The cats had been seized in a raid of an animal sanctuary that was not caring for them properly. This required hundreds of cats to be spayed or neutered, and adopted out to new owners. The cats are prepped for surgery and anesthetized, and what is pictured is apparently is pretty standard procedure. The cats are not being experimented upon – they are asleep and under medical care.
You can read more about how image search works at Google’s help page here. The right click option “Search Google for this image” is built in to current versions of the Chrome browser. But it can be added to other browsers like Firefox via various add-ons or plugins.
It is worth noting that photos that go wildly viral will often poison the image search results – all you’ll find at the top is more viral posts on various Twitter feeds, Tumblrs and blogs. But in typical cases, the original post is older – sometimes quite a bit older. (This post was viral in 2013 and 2014, but the original photo was from 2011). So we can use date functions in the search engine to solve this problem.
TinEye now has a sort by oldest option, prominently marked in the left sidebar, as pictured at left. That will move the picture you want to check first right up to the top of the results.
Google Image Search does this via the buttons above the results. Click “Search tools” and then the “Time” drop down which appears. Pick “Custom range…” to choose dates. Here’s what it looks like:
For your custom range, pick a start date of many years ago, and an end date before you think the viral may have started. If you still see similar viral posts, keep backing up the date at the end of the range. Once the viral posts disappear from the results, what is left may be the original source of the photo.
There you go, a super-simple tool (that may already be in your browser) that can help you be more skeptical before retweeting the next viral sensation.
And remember, the time you need to do this most is when the content in front of you feeds into your own confirmation bias about your personal passions, whatever those are.