Bloch’s Mirror: how a 1940 defeat illuminates modern cognitive warfare

In June 2026, the French his­tor­i­an Marc Bloch will be induc­ted into the Panthéon. Today, his work is read primar­ily for its eth­ic­al dimen­sion, and while this inter­pret­a­tion is val­id, it is argu­ably reduct­ive. The Strange Defeat, his account of the 1940 debacle, is also a rig­or­ous obser­va­tion of a mil­it­ary sys­tem that loses the capa­city to pro­duce a shared oper­a­tion­al real­ity. Re-examined today through the lens of cog­nit­ive sci­ence, this text offers a case study of decision-mak­ing fail­ure under inform­a­tion overload.

This new inter­pret­a­tion is not attrib­ut­ing a the­ory to Bloch that he did not put for­ward; rather it uses con­tem­por­ary ana­lyt­ic­al tools to cla­ri­fy what he observed. The hypo­thes­is is simple: the mech­an­isms he described in 1940 cor­res­pond to the vul­ner­ab­il­it­ies we see today in mod­ern com­mand sys­tems. With­in this frame­work, a think tank pro­ject recently con­duc­ted over sev­er­al months with the Chair of the Centre for High­er Mil­it­ary Edu­ca­tion (CEMST) com­bined an ana­lys­is of Bloch’s body of work, con­cep­tu­al form­al­isa­tion (dis­trib­uted cog­ni­tion, sense-mak­ing, decision-mak­ing cycles) and a com­par­is­on with data from recent com­mand and con­trol (C2) exercises.

The French army in 1940 was not ill-equipped. It fielded nearly 2.5 to 3 mil­lion men, some 3,000 tanks and a sub­stan­tial artil­lery force. The dif­fer­ence in equip­ment com­pared with the Wehr­macht does not, on its own, explain the col­lapse with­in six weeks. What Bloch high­lights is of a dif­fer­ent nature. The prob­lem was not a lack of inform­a­tion, but an inab­il­ity to integ­rate it. Inform­a­tion flows cir­cu­lated without con­ver­ging. Assess­ments differed from one level of com­mand to the next. Decisions arrived too late to have any effect. Sev­er­al “real­it­ies” coex­is­ted with­in the same sys­tem, with no reli­able mech­an­ism to recon­cile them.

Con­tem­por­ary research on digit­ised com­mand posts con­firms this dynam­ic. Bey­ond a cer­tain threshold of inform­a­tion dens­ity (in the order of 10³ to 10⁴ events per hour depend­ing on the con­fig­ur­a­tion), the coher­ence of shared rep­res­ent­a­tions drops by 30 to 50 per cent. It is not a lack of data that causes the error: it is their fragmentation.

Two pro­cesses under­pin the break­down described by Bloch:

The rigid­ity of inter­pret­at­ive frame­works. French officers inter­preted 1940 through the frame­work of events from 1914–1918. New inform­a­tion was absorbed into obsol­ete cat­egor­ies, res­ult­ing in a sys­tem­at­ic­ally flawed ori­ent­a­tion. The soci­olo­gist Karl Weick form­al­ised this pro­cess under the term “sense­mak­ing”: an organ­isa­tion acts based on estab­lished inter­pret­a­tions, not raw facts. When these frame­works become inad­equate, cor­rec­tion is slow. In recent joint exer­cises, 35 to 45 per cent of ini­tial decisions rep­lic­ate doc­trin­al frame­works ill-suited to non-stand­ard situations.

Inform­a­tion over­load. Bloch emphas­ises the abund­ance of inform­a­tion received and its lim­ited oper­a­tion­al value. This is what Her­bert Simon con­cep­tu­al­ised: an excess of inform­a­tion cre­ates a scarcity of atten­tion. Bey­ond a cer­tain threshold, crit­ic­al sig­nals are drowned out in the flow, pri­or­it­isa­tion breaks down, and errors increase dis­pro­por­tion­ately. Exper­i­ment­al stud­ies meas­ure per­form­ance declines of 20 to 50 per cent when inform­a­tion flows are unfiltered.

These two mech­an­isms rein­force each oth­er. Inter­pret­at­ive rigid­ity pre­vents the recon­fig­ur­a­tion of pri­or­it­ies; sat­ur­a­tion pre­vents the iden­ti­fic­a­tion of rel­ev­ant sig­nals. The sys­tem con­tin­ues to “func­tion”, but it no longer pro­duces coher­ent decisions.

Cur­rent ana­lyt­ic­al frame­works enable us to cla­ri­fy the dynam­ics that Bloch observed. Oper­a­tion­al cog­ni­tion is dis­trib­uted, in the sense described by the research­er Edwin Hutchins, which is that decision-mak­ing is not the work of a single indi­vidu­al, it rather emerges from the inter­ac­tions between play­ers, tools and pro­ced­ures. In this sys­tem, the integ­ra­tion phase, or “ori­ent­a­tion” in mil­it­ary the­or­ist John Boyd’s mod­el, accounts for the bulk of the vul­ner­ab­il­ity – it rep­res­ents up to 65–75% of the decision-mak­ing cycle time, and it is here that ambi­gu­ity and over­load cause the most damage.

The work­ing group with the CEMST Chair has trans­lated this vul­ner­ab­il­ity into meas­ur­able indic­at­ors. Three cat­egor­ies have been iden­ti­fied: con­sist­ency of rep­res­ent­a­tions (degree of con­ver­gence between com­mand cells), decision-mak­ing latency (time between receiv­ing crit­ic­al inform­a­tion and the decision), and inter­pret­at­ive sta­bil­ity (vari­ab­il­ity of assess­ments over short intervals). 

When inter-cell coher­ence falls by more than 25–30%, decision-mak­ing latency increases by 20 to 40%

When applied to real C2 exer­cises, these indic­at­ors reveal robust cor­rel­a­tions. When inter-cell coher­ence falls by more than 25–30%, decision-mak­ing latency increases by 20 to 40%. Bey­ond 40% diver­gence, con­tra­dict­ory decisions coex­ist sim­ul­tan­eously with­in the sys­tem, with imme­di­ate neg­at­ive oper­a­tion­al effects. 

The tem­por­al dimen­sion is decis­ive here. A 10–20% advant­age in the decision-mak­ing cycle is suf­fi­cient, in many scen­ari­os, to pro­duce a decis­ive tac­tic­al advant­age. Bloch describes a sys­tem that is struc­tur­ally behind sched­ule, not due to a lack of inform­a­tion, but due to an inab­il­ity to trans­form it into a shared direction. 

The par­al­lel with Gra­ham Allison’s work is reveal­ing. In Essence of Decision (1971), Allis­on shows that stra­tegic decision-mak­ing is con­strained by organ­isa­tion­al routines and bur­eau­crat­ic logic. Bloch observes a sim­il­ar phe­nomen­on in real, large-scale war­time situ­ations. Where­as Allis­on mod­els his ana­lys­is based on the Cuban Mis­sile Crisis, Bloch draws on evid­ence from the front lines of 1940. The con­ver­gence does not stem from a lin­eage but from a shared struc­ture: the fra­gil­ity of col­lect­ive decision-mak­ing under pressure.

These find­ings have dir­ect prac­tic­al implic­a­tions. They argue for inform­a­tion frame­works that lim­it frag­ment­a­tion (hier­arch­ic­al chan­nels, struc­tured aggreg­a­tion), fil­ter­ing mech­an­isms that reduce inform­a­tion over­load (pri­or­it­isa­tion, alert thresholds), and train­ing aimed at fos­ter­ing cog­nit­ive flex­ib­il­ity. That is to say, the abil­ity to rap­idly revise one’s inter­pret­at­ive frame­works. They also sup­port the integ­ra­tion of con­sist­ency and latency met­rics into the eval­u­ation of com­mand staff.

This rein­ter­pret­a­tion does not make Bloch a the­or­ist of cog­nit­ive war­fare. It estab­lishes a pre­cise cor­res­pond­ence between a his­tor­ic­al obser­va­tion and con­tem­por­ary ana­lyt­ic­al frame­works. The Strange Defeat thus emerges as an empir­ic­al case of a fail­ure of decision-mak­ing coher­ence in a sat­ur­ated envir­on­ment. Cur­rent research does not “con­firm” Bloch in the his­tor­ic­al sense; rather, it enables us to spe­cify, meas­ure and com­pare the mech­an­isms he described.

In envir­on­ments where inform­a­tion dens­ity and the speed of engage­ment are con­stantly increas­ing, the cent­ral issue is no longer merely a mat­ter of hav­ing inform­a­tion. It is a mat­ter of pro­du­cing, with­in the time avail­able for action, a shared oper­a­tion­al real­ity. It is pre­cisely this capa­city that Bloch shows was deteri­or­at­ing in 1940, and which the work car­ried out at the CEMST now seeks to char­ac­ter­ise and preserve.