Background: The Journal

Some­times it is easy to lose sight of still-es­sen­tial as­pects of the schol­arly journ­al, es­pe­cially when con­tem­plat­ing rad­ic­al re­forms of the sys­tem. Aim­ing to foster a more well-roun­ded pic­ture, we re­view here a few as­pects of the journ­al, its roles and con­text.

Functional View

Legacy functions

  • dis­sem­in­a­tion;
  • re­gis­tra­tion (= re­cord­ing who dis­covered first);
  • archiv­ing (= es­tab­lish­ing the ver­sion of re­cord).

These func­tions were quite cent­ral in the past to the pur­pose of the journ­al; nowadays, they can also be eas­ily achieved through oth­er means.

Auxiliary functions

  • pre­par­a­tion (= re­vi­sion after re­views, then copy­ed­it­ing/type­set­ting);
  • defin­i­tion (= out­lining a sub­field and a com­munity around it).

These two func­tions are still quite im­port­ant (in­deed, the pro­cess of re­vis­ing upon feed­back from re­view­ers could be seen as a con­tinu­ation, some­times even a prop­er part, of the re­search it­self), but they could (at least con­ceiv­ably) be served through mech­an­isms oth­er than the journ­al.

Core functions

  • val­id­a­tion (= the res­ult is sound);
  • fil­tra­tion (= helps se­lect what’s worth read­ing to whom);
  • sig­nal­ing (= helps se­lect who’s worth pro­mot­ing or fund­ing).

To our mind, it’s these three func­tions, knot­ted in a single so­cial in­sti­tu­tion, that make the journ­al still in­dis­pens­able today.

Transactional View

Predictive channel

  • The au­thor gets at­ten­tion: the work is seen by the right audi­ence of peers;
  • The read­er gets sources: she re­ceives se­lec­ted and sound res­ults on which to build re­search.

The journ­al me­di­ates bring­ing art­icles to read­ers, stamped by a stat­ist­ic­al meas­ure (the journ­al’s pro­file and renown, built mostly as an av­er­age of past art­icles) that they’re worth spend­ing the time to read. That this meas­ure is im­per­fect does not make it use­less, and is bet­ter than oth­er short­cuts read­ers might take to pro­tect their time.

Evaluative channel

  • The au­thor gets prestige, as the journ­al’s renown re­flects on the au­thor.
  • The fun­der/man­ager gets eval­u­ations: a meas­ure­ment of an au­thor’s work, again through the renown of the journ­al.

The pre­dict­ive and eval­u­at­ive sides could in prin­ciple be de­coupled (e.g., per­formed by dif­fer­ent ser­vices). In­deed, DORA (= the San Fran­cisco De­clar­a­tion on Re­search As­sess­ment, 2012) calls for eval­u­ation of re­search to be done in­de­pend­ent of the Journ­al Im­pact Factor (a skewed nu­mer­ic re­place­ment for the journ­al’s pro­file), and fur­ther, in­de­pend­ent of an art­icle’s journ­al ven­ue al­to­geth­er. However, this leaves un­touched the need for the pre­dict­ive side (used for fil­ter­ing be­fore any thor­ough read­ing).

diagram of a journal as multi-sided

An eco­nom­ics view would be to mod­el each journ­al as a multi-sided mar­ket (a plat­form). Worth no­ti­cing are the read­ers’ seek­ing costs (the ef­fort and time spent find­ing an art­icle that would re­ward close read­ing), and that a sub­scrip­tion does not pur­chase art­icles as such, but an op­tion value (the avail­ab­il­ity of a stream of art­icles, on the prob­ab­il­ity of need­ing one of them in the fu­ture).

Schol­arly com­mu­nic­a­tion is built on small net­works: each au­thor is con­nect­ing with re­l­at­ively few read­ers. (Com­par­is­ons with iTunes, tele­vi­sion, or oth­er me­dia at mass-mar­ket scales may be tempt­ing, but are flawed.)

Social Institution

The journ­al is an es­tab­lished so­cial in­sti­tu­tion, and while it settled in its cur­rent form in the age of print, the com­munity’s habit of par­ti­cip­at­ing in the com­mons of schol­arly know­ledge, as ed­it­ors and re­view­ers, has sur­vived. For ex­ample, re­view­ers are not dir­ectly paid in part be­cause re­view­ing is un­der­stood as part of be­ing a schol­ar and help­ing the schol­arly com­munity.

Such shared un­der­stand­ings are easi­er to des­troy than launch or re­place. (The lack of suc­cess with post-pub­lic­a­tion peer-re­view mod­els — e.g., for­um-like com­ments on pub­lished ma­ter­i­als, with con­tinu­al up­dates, cor­rec­tions, eval­u­ations — is one ex­ample of a nice the­ory not mov­ing much of the prac­tice.) Ima­gined changes in col­lect­ive ac­tion are pain­less to pos­tu­late, but hard to es­tab­lish as func­tion­ing in­sti­tu­tions. If you merely build it, of­ten nobody will come.


Format­ting is a key func­tion of schol­arly com­mu­nic­a­tion, as the present­a­tion and lay­out of art­icles as­sists both fil­ter­ing and as­sim­il­a­tion. The pat­terns and con­ven­tions of the field (e.g., defin­i­tion–the­or­em–proof; in gen­er­al, the typ­ic­al struc­ture of an art­icle) of­fer guid­ance and can ac­cel­er­ate the read­er; poor writ­ing or ty­po­graphy poses obstacles. (The sens­it­iv­ity to these obstacles seems to dif­fer greatly among read­ers.)

The sur­pris­ing sur­viv­al of PDF as the format of choice (across all schol­arly pub­lish­ing) hints that the lay­out tra­di­tions from the age of print still re­main the most com­fort­able, read­able or un­der­stand­able, and again, that format­ting is im­port­ant.

A Rising Pressure

Across all schol­arly fields, vari­ous growth fig­ures are re­por­ted: the num­ber of journ­als is in­creas­ing at ~3% per year; the num­ber of re­search­ers, at ~4%; the amount spent on R&D, at ~3% (on top of in­fla­tion); the num­ber of cite­able items, at ~8% (doubles every nine years). The ex­act num­bers mat­ter less than the trend and the pres­sure ad­ded: It be­comes harder to find what you need to read; it be­comes harder to eval­u­ate someone’s re­search out­put. An in­creas­ing abund­ance of con­tent leads to an in­creas­ing scarcity of re­search­ers’ time and at­ten­tion.

This makes the core func­tions of the journ­al more, not less, ne­ces­sary. The con­com­it­ant need for more fine-grained fil­ter­ing, rather than al­low that journ­als be­come su­per­flu­ous (in fa­vor of, for ex­ample, more amorph­ous pub­lic­a­tion on pre­print serv­ers or after sound­ness-only re­view), in­stead leads to fur­ther in­creases in the num­ber and gran­u­lar­ity of journ­als, at least for the near fu­ture.

To our minds, the im­port­ance of the schol­arly journ­al could di­min­ish if its roles were to be sup­planted, for ex­ample by large data-driv­en smart ser­vices, provid­ing finer clas­si­fic­a­tion and match­ing. However, such ser­vices would most likely be­come both in­dis­pens­able and ex­pens­ive, and might well end up hold­ing con­trolling power over schol­arly com­mu­nic­a­tion and ca­reers, more than tra­di­tion­al pub­lish­ing ever did. Fos­ter­ing a healthy and vi­brant eco­logy of schol­ar-led journ­als might provide a coun­ter­bal­ance and should be prefer­able.

Pos­ted in May 2019

Testimonials from Authors  
MSP’s thoughts and approach to open access  
Policies for Authors