The rela­tions between thought, lan­guage, and mood have become sub­jects of study for sev­er­al sci­en­tif­ic fields of late. Some of the con­clu­sions seem to echo reli­gious notions from mil­len­nia ago. “As a man thin­keth, so he is,” for exam­ple, pro­claims a famous verse in Proverbs (one that helped spawn a self-help move­ment in 1903). Pos­i­tive psy­chol­o­gy might agree. “All that we are is the result of what we have thought,” says one trans­la­tion of the Bud­dhist Dhamma­pa­da, a sen­ti­ment that cog­ni­tive behav­ioral ther­a­py might endorse.

But the insights of these tra­di­tions — and of social psy­chol­o­gy — also show that we’re embed­ded in webs of con­nec­tion: we don’t only think alone; we think — and talk and write and read — with oth­ers. Exter­nal cir­cum­stances influ­ence mood as well as inter­nal states of mind. Approach­ing these ques­tions dif­fer­ent­ly, researchers at the Lud­dy School of Infor­mat­ics, Com­put­ing, and Engi­neer­ing at Indi­ana Uni­ver­si­ty asked, “Can entire soci­eties become more or less depressed over time?,” and is it pos­si­ble to read col­lec­tive changes in mood in the writ­ten lan­guages of the past cen­tu­ry or so?

The team of sci­en­tists, led by Johan Bollen, Indi­ana Uni­ver­si­ty pro­fes­sor of infor­mat­ics and com­put­ing, took a nov­el approach that brings togeth­er tools from at least two fields: large-scale data analy­sis and cog­ni­tive-behav­ioral ther­a­py (CBT). Since diag­nos­tic cri­te­ria for mea­sur­ing depres­sion have only been around for the past 40 years, the ques­tion seemed to resist lon­gi­tu­di­nal study. But CBT pro­vid­ed a means of ana­lyz­ing lan­guage for mark­ers of “cog­ni­tive dis­tor­tions” — think­ing that skews in over­ly neg­a­tive ways. “Lan­guage is close­ly inter­twined with this dynam­ic” of thought and mood, the researchers write in their study, “His­tor­i­cal lan­guage records reveal a surge of cog­ni­tive dis­tor­tions in recent decades,” pub­lished just last month in PNAS.

Choos­ing three lan­guages, Eng­lish (US), Ger­man, and Span­ish, the team looked for “short sequences of one to five words (n‑grams), labeled cog­ni­tive dis­tor­tion schema­ta (CDS).” These words and phras­es express neg­a­tive thought process­es like “cat­a­stro­phiz­ing,” “dichoto­mous rea­son­ing,” “dis­qual­i­fy­ing the pos­i­tive,” etc. Then, the researchers iden­ti­fied the preva­lence of such lan­guage in a col­lec­tion of over 14 mil­lion books pub­lished between 1855 and 2019 and uploaded to Google Books. The study con­trolled for lan­guage and syn­tax changes dur­ing that time and account­ed for the increase in tech­ni­cal and non-fic­tion books pub­lished (though it did not dis­tin­guish between lit­er­ary gen­res).

What the sci­en­tists found in all three lan­guages was a dis­tinc­tive “‘hock­ey stick’ pat­tern” — a sharp uptick in the lan­guage of depres­sion after 1980 and into the present time. The only spikes that come close on the time­line occur in Eng­lish lan­guage books dur­ing the Gild­ed Age and books pub­lished in Ger­man dur­ing and imme­di­ate­ly after World War II. (High­ly inter­est­ing, if unsur­pris­ing, find­ings.) Why the sud­den, steep climb in lan­guage sig­ni­fy­ing depres­sive think­ing? Does it actu­al­ly mark a col­lec­tive shift in mood, or show how his­tor­i­cal­ly oppressed groups have had more access to pub­lish­ing in the past forty years, and have expressed less sat­is­fac­tion with the sta­tus quo?

While they are care­ful to empha­size that they “make no causal claims” in the study, the researchers have some ideas about what’s hap­pened, observ­ing for exam­ple:

The US surge in CDS preva­lence coin­cides with the late 1970s when wages stopped track­ing increas­ing work pro­duc­tiv­i­ty. This trend was asso­ci­at­ed with ris­es in income inequal­i­ty to recent lev­els not seen since the 1930s. This phe­nom­e­non has been observed for most devel­oped economies, includ­ing Ger­many, Spain and Latin Amer­i­ca.

Oth­er fac­tors cit­ed include the devel­op­ment of the World Wide Web and its facil­i­ta­tion of polit­i­cal polar­iza­tion, “in par­tic­u­lar us-vs.-them think­ing… dichoto­mous rea­son­ing,” and oth­er mal­adap­tive thought pat­terns that accom­pa­ny depres­sion. The scale of these devel­op­ments might be enough to explain a major col­lec­tive rise in depres­sion, but one com­menter offers an addi­tion­al gloss:

The globe is *Lit­er­al­ly* on fire, or his­tor­i­cal­ly flood­ing — Mul­ti­ple eco­nom­ic crash­es bare­ly decades apart — a ghost town of a hous­ing mar­ket — a mul­ti-year glob­al pan­dem­ic — wealth con­cen­tra­tion at the .01% lev­el — ter­ri­ble pay/COL equa­tions — block­ing unionization/workers rights — abu­sive mil­i­ta­rized police, with­out the restraint or train­ing of actu­al mil­i­tary —  You can’t afford X for a month­ly mort­gage pay­ment!  Pay 1.5x for rent instead! — end­less wars for the last… 30…years? 50 if we include stuff like Korea, Cold War, Viet­nam… How far has the IMC been milk­ing the gov for funds to make the rich rich­er? Oh, and a bil­lion­aire 3‑way space race to deter­mine who’s got the biggest “rock­et”

These sound like rea­sons for glob­al depres­sion indeed, but the arrow could also go the oth­er way: maybe cat­a­stroph­ic rea­son­ing pro­duced actu­al cat­a­stro­phes; black and white think­ing led to end­less wars, etc…. More study is need­ed, says Bollen and his col­leagues, yet it seems prob­a­ble, giv­en the data, that “large pop­u­la­tions are increas­ing­ly stressed by per­va­sive cul­tur­al, eco­nom­ic, and social changes” — changes occur­ring more rapid­ly, fre­quent­ly, and with greater impact on our dai­ly lives than ever before. Read the full study at PNAS. 

Relat­ed Con­tent: 

Stanford’s Robert Sapol­sky Demys­ti­fies Depres­sion, Which, Like Dia­betes, Is Root­ed in Biol­o­gy

A Uni­fied The­o­ry of Men­tal Ill­ness: How Every­thing from Addic­tion to Depres­sion Can Be Explained by the Con­cept of “Cap­ture”

Charles Bukows­ki Explains How to Beat Depres­sion: Spend 3–4 Days in Bed and You’ll Get the Juices Flow­ing Again (NSFW)

Josh Jones is a writer and musi­cian based in Durham, NC. Fol­low him at @jdmagness