Louder or longer? Optimizing information transmission in animals and the evolution of language

A new study pub­lished in Evol­u­tion Let­ters sug­gests vocal rep­er­toire effi­ciency in rock hyraxes is driv­en by call amp­litude rather than dur­a­tion, chal­len­ging pre­dic­tions of the Law of Brev­ity. Authors Dr Vlad Demart­sev and Dr. Amyi­aal Ilany tell us more.

Social liv­ing in nature is tightly bound with the abil­ity to com­mu­nic­ate with oth­ers. Main­tain­ing social ties and coordin­at­ing with group mates require fre­quent com­mu­nic­a­tion; there­fore, com­plex social sys­tems are usu­ally asso­ci­ated with well-developed com­mu­nic­a­tion abil­it­ies. The apex of com­mu­nic­a­tion­al com­plex­ity is undoubtedly human lan­guage; char­ac­ter­ised by a nearly unlim­ited capa­city for inform­a­tion trans­mis­sion, it is con­sidered to be one of the main factors facil­it­at­ing the form­a­tion of com­plex human soci­ety and cul­ture. How­ever, intens­ive and inform­a­tion­ally rich com­mu­nic­a­tion comes at a cost in terms of time spent trans­mit­ting the inform­a­tion and of mus­cu­lar effort, inves­ted to artic­u­late the sig­nals. Thus, sig­nalling sys­tems are likely to become optim­ized by min­im­iz­ing efforts for sig­nalling unit pro­duc­tion, while main­tain­ing its inform­a­tion­al con­tent.  Amer­ic­an lin­guist George Kings­ley Zipf has pop­ular­ized this concept in the 1930s by artic­u­lat­ing the Law of Brev­ity, stat­ing that a word length is neg­at­ively cor­rel­ated with its fre­quency of use in a lan­guage, i.e. the most fre­quently used words are short­er. This prin­ciple was veri­fied in almost a thou­sand lan­guages and it is reg­u­larly observed in the pro­cess of lan­guage evol­u­tion. Spar­ing their effort, speak­ers tend to shorten long words, when they become more fre­quently used; e.g. the word “tele­vi­sion” has been shortened to TV.

The uni­ver­sal nature of the Law of Brev­ity raised the ques­tion of wheth­er it is a gen­er­al char­ac­ter­ist­ic of any com­mu­nic­a­tion­al sys­tem and wheth­er its exist­ence in human lan­guages stems from the evol­u­tion­ary ori­gins of anim­al com­mu­nic­a­tion. The rela­tion­ship between call dur­a­tion and usage fre­quency was tested in sev­er­al anim­als, but the res­ults differed between the spe­cies. A pro­posed explan­a­tion for the lack of a clear fit of anim­al rep­er­toires to the brev­ity prin­ciple is the abund­ance of long-range calls. Humans mostly com­mu­nic­ate from a short-range (< 3.5 meters) while anim­als fre­quently need to trans­mit their sig­nals much fur­ther. Longer calls are more effi­cient for long-range com­mu­nic­a­tion, as they are less likely to be masked by noise. This might have res­ul­ted in a con­tra­dict­ing pres­sure against short­en­ing of vocal sig­nals. Addi­tion­ally, long-range calls need to be louder, which is likely adding to their pro­duc­tion effort.

Fol­low­ing this dir­ec­tion, we wanted to exam­ine wheth­er call amp­litude, rather than call dur­a­tion, might be the main factor by which anim­al vocal rep­er­toires are optim­ized. By adopt­ing the “least-effort” logic, i.e. fre­quent calls should require least effort to be pro­duced, we have hypo­thes­ized that softer calls would be more fre­quent than louder ones.

We tested this in rock hyraxes, a medi­um-sized mam­mal nat­ive to Africa and the Middle East. Rock hyraxes are social anim­als, liv­ing in groups of up to 20 indi­vidu­als. The groups con­sist of mul­tiple females and their off­spring with usu­ally one adult res­id­ent male. With­in the group, hyraxes fre­quently com­mu­nic­ate, using an extens­ive rep­er­toire of calls.  On the oth­er hand, bach­el­or adult males (mostly lead­ing sol­it­ary lives) inter­act with females only briefly dur­ing the short mat­ing sea­son and have mainly aggress­ive encoun­ters with oth­er males. Males fre­quently sing com­plex and loud advert­ise­ment songs, trans­mit­ting their indi­vidu­al qual­ity to both females and neigh­bour­ing males.

Our hyrax study pop­u­la­tion is resid­ing at the Ein-Gedi Nat­ur­al Reserve near the Dead Sea in Israel. This wild pop­u­la­tion has been con­tinu­ously mon­itored since 1999 as a part of a long term study of hyrax beha­viour, com­mu­nic­a­tion and social­ity, led by Prof. Eli Gef­fen (School of Zoology, Tel-Aviv Uni­ver­sity) and recently also by Dr. Amyi­aal Ilany and Dr. Lee Koren (The Mina and Ever­ard Good­man Fac­ulty of Life Sci­ences, Bar-Ilan Uni­ver­sity). As part of this study, we have fit­ted 19 hyraxes (males and females) with indi­vidu­al, mini­ature audio record­ers and logged all their calls over approx­im­ately one week. By listen­ing and labelling all recor­ded calls, we cre­ated full rock hyrax vocal rep­er­toire. Using this extens­ive data­set, we have cal­cu­lated the usage fre­quen­cies of all call types and have meas­ured the aver­age dur­a­tion and amp­litude for each one. This allowed us to exam­ine if hyrax vocal rep­er­toire cor­res­ponds with the clas­sic Law of Brev­ity (call duration/usage) rela­tion­ship, or, wheth­er the optim­iz­a­tion factor of the vocal per­form­ance is call amplitude.

When com­par­ing male and female rep­er­toires, we found that females pro­duce more call types in gen­er­al and more affil­i­at­ive call types in par­tic­u­lar. This was not sur­pris­ing, as hyrax females main­tain stable social rela­tion­ships with­in a group, while bach­el­or males have only lim­ited com­mu­nic­a­tion oppor­tun­it­ies.  There are sexu­al dif­fer­ences also in rela­tion to the fit to Law of Brev­ity. In females, longer calls are actu­ally the more fre­quent ones, in con­tra­dic­tion to the Law of Brevity’s pre­dic­tion. In con­trast, amp­litude seems to fol­low the “least effort” paradigm, as soft calls (requir­ing less effort to pro­duce) are more fre­quently used.  The male rep­er­toire, on the oth­er hand, is optim­ised for min­im­iz­ing dur­a­tion, but not amp­litude. Male vocal­iz­a­tions are heav­ily influ­enced by the unique require­ments of their advert­ise­ment songs, which have to be loud in order to reach remote listen­ers. There­fore, optim­iz­a­tion of male rep­er­toire for amp­litude would have likely impaired the songs func­tion. When songs are excluded from the ana­lys­is, amp­litude-based optim­iz­a­tion is detec­ted for the remain­ing male repertoire.

In hyraxes, the two sexes were under dif­fer­ent select­ive pres­sures in terms of rep­er­toire optim­iz­a­tion. Female rep­er­toire, tuned for fre­quent com­mu­nic­a­tion over a wide range of dis­tances, is optim­ized by min­im­iz­ing call amp­litude. How­ever, while amp­litude is the main can­did­ate for optim­iz­a­tion, this can be con­strained by con­tra­dict­ing selec­tion pres­sures. Males are com­mit­ted to their loud songs, so amp­litude reduc­tion is pre­ven­ted by the needs for long-range trans­mis­sion, res­ult­ing only in dur­a­tion-based optim­iz­a­tion of their rep­er­toire, much like in human lan­guages.  It is thus an intriguing ques­tion; why aren’t human lan­guages optim­ized by amp­litude? Could it be because the devel­op­ment of arti­fi­cial sig­nalling means for long-range com­mu­nic­a­tion made high amp­litude calls less needed? Per­haps the high pres­sure for increased inform­a­tion­al con­tent in the emer­ging human lan­guages capped the amp­litude of the vocal sig­nals, as loud calls have less capa­city for inform­a­tion­al con­tent. Both scen­ari­os could lead to dur­a­tion-based optim­iz­a­tion that is now widespread.

Dr. Amyi­aal Ilany is a seni­or lec­turer at the Fac­ulty of Life Sci­ences, Bar Ilan Uni­ver­sity. The ori­gin­al study is freely avail­able to read and down­load from Evol­u­tion Let­ters.