Electronic Data Processing (EDP) in Traffic Matters (Lecture)

Electronic Data Processing (EDP) in Traffic Matters


Three questions at the beginning

Let me start with three provoca­tive questions:

- Is there any rep­re­sen­ta­tive group of lawyers in Ger­many, which is real­ly enthusiastic
about com­put­ers in traf­fic mat­ters or court cas­es in general?

- How fre­quent­ly do we come across the opin­ion that EDP should be implemented
as a mat­ter of prin­ci­ple rather than mere­ly for prac­ti­cal purposes?

- Is it not the case that EDP has been met with a rather reluc­tant acceptance?

Every­one will cer­tain­ly have their own answers to such ques­tion. These answers form the crux of the mat­ter when inves­ti­gat­ing the top­ic of EDP in Traf­fic Mat­ters, and it is nec­es­sary to exam­ine them carefully.

Potential and Necessitated Applications

As is the case with any reflec­tion on tech­ni­cal mat­ters, it is worth­while from the out­set to con­sid­er method­i­cal­ly its poten­tial appli­ca­tions, as well as the instances that neces­si­tate its application.

With­in the broad range of poten­tial appli­ca­tions avail­able there is one aspect of great impor­tance: We should only seize upon an oppor­tu­ni­ty if there is good rea­son for our doing so. If such rea­sons exist, the sit­u­a­tion may even devel­op into a neces­si­ty. In oth­er words (using “neces­si­ty” in its lit­er­al sense): if the need dictates.

Do we face a neces­si­ty when it comes to traf­fic mat­ters? Many of those who are involved in these mat­ters are aware of the urgency of the sit­u­a­tion, a sit­u­a­tion for which they them­selves are blame­less. How­ev­er, the mul­ti­tude of cas­es flood­ing the jus­tice sys­tem and the back­log which ensues has reached alarm­ing dimen­sions. Using the term “neces­si­ty” in this con­text may thus be no exag­ger­a­tion. Hold-ups and tail­backs on the judi­cial high­way, if we are to stick to traf­fic-relat­ed ter­mi­nol­o­gy, are haz­ards that we must con­front, espe­cial­ly if we want to secure the integri­ty of our judi­cial system.

Method of Approach

How can we approach our ques­tion method­i­cal­ly and pur­pose­ful­ly? My approach is inspired by the French philoso­pher Roland Barthes, mem­ber of the Coll�ge de France. In his life­long philo­soph­i­cal rea­son­ing he paid par­tic­u­lar atten­tion to traf­fic mat­ters. He was firm­ly con­vinced that the world is a sys­tem of signs which can be read like a book. In this sense, traf­fic is an espe­cial­ly mean­ing­ful book.

Main Characteristics of the Phenomenon of “Traffic”

Let us first take into con­sid­er­a­tion the struc­ture of traf­fic and there­fore imple­ment a phe­nom­e­no­log­i­cal approach. What do we under­stand by the terms traf­fic and traf­fic matters?

Traf­fic mat­ters are mat­ters con­cern­ing traf­fic. This is obvi­ous­ly the expres­sion put sim­ply. From this approach (an approach which I con­sid­er to be rea­son­able), traf­fic mat­ters are sub­ject to def­i­n­i­tion as a kind of epiphe­nom­e­non and thus, pre­sum­ably, they share cer­tain cen­tral char­ac­ter­is­tics with traf­fic. Let us, there­fore, turn our atten­tion first to the phe­nom­e­non of traf­fic. What char­ac­ter­is­tics imme­di­ate­ly strike us when we exam­ine traf­fic from a distance?

Traffic’s enormity

To begin with, there is traffic’s enor­mi­ty. Traf­fic is a mas­sive phe­nom­e­non. Even pri­vate vehi­cle traf­fic, a term fre­quent­ly mis­con­strued, is a mass phe­nom­e­non, and this is where the prob­lem lies.


Sec­ond­ly, there is uni­for­mi­ty. When regard­ed from the per­spec­tive just men­tioned, indi­vid­ual car brands lose their impor­tance. All vehi­cles, be they lux­u­ry or com­pact cars, blend in an abstract sense into one stream of traf­fic mov­ing past our ana­lyt­ic eye.


And third­ly, there is speed, which is gen­er­al­ly tak­en to be cru­cial to traf­fic, but although it often can­not be fac­tu­al­ly com­pre­hend­ed, it exists as a “will and a vision”. The term “stand­still” seems para­dox­i­cal, as the wish to main­tain an appro­pri­ate speed lev­el in order to cov­er dis­tances is cru­cial to traffic.

Central Characteristics of Traffic Matters

Traf­fic mat­ters share the char­ac­ter­is­tics of enor­mi­ty, uni­for­mi­ty and speed, which them­selves are respon­si­ble for hav­ing passed on these fea­tures to them in the first place and thus put their seal on them.


Our first encounter is with mass phe­nom­e­na. As we all know, the cas­es relat­ing to traf­fic have reached stag­ger­ing pro­por­tions. It is often dif­fi­cult for our legal sys­tem to cope with the flood of cas­es, and in many places, the lim­its of cop­ing with this mul­ti­tude have reached, if not exceeded.


Sec­ond­ly, we can see that uni­for­mi­ty, too, exerts its influ­ence in this con­text. Traf­fic relat­ed mat­ters demon­strate uni­for­mi­ty to a much greater extent than oth­er domains. Typ­i­cal inci­dents keep recur­ring. Cen­tral to these is the road traf­fic acci­dent, and this gives traf­fic mat­ters their indi­vid­ual character.


Third­ly, we encounter speed in traf­fic mat­ters in the same man­ner as has been described for traf­fic and this needs to be tak­en into consideration.

The Promises of EDP:
Achieving more, more Uniformity and improving Speed

Advo­cates of com­put­er­i­sa­tion are par­tic­u­lar­ly pleased when we name the three com­pa­ra­ble cen­tral char­ac­ter­is­tics shared by traf­fic and traf­fic mat­ters: enor­mi­ty, uni­for­mi­ty and speed; and they say: This instance is suit­ed the appli­ca­tion of EDP. The use of EDP is ide­al­ly suit­ed to the man­age­ment of mass occur­rences. It ensures uni­form and quick han­dling of mat­ters. Elec­tron­ic data pro­cess­ing helps us to tack­le prob­lems in a more effec­tive and homo­ge­neous way.
This is the promise of elec­tron­ic data pro­cess­ing (EDP) which we encounter every­where, and this is why any orga­ni­za­tion con­cerned with legal mat­ters is already equipped with computers.

Intermediate Considerations

At this point we must tread more cau­tious­ly as one of the risks has already been real­ized. If we look more close­ly at what has hap­pened: a fac­tu­al sit­u­a­tion aris­es, a promise is giv­en and enor­mous invest­ment ensues. But is the appli­ca­tion of EDP in this field jus­ti­fied? What makes us trust the promis­es made by EDP? When we ask this ques­tion the world changes. That is why we should focus on this basic prin­ci­ple more fre­quent­ly than we usu­al­ly do. The essence of this prin­ci­ple is easy to find: Needs do not arise auto­mat­i­cal­ly. Only when an appli­ca­tion is held against nor­ma­tive pre­req­ui­sites which them­selves demand jus­ti­fi­ca­tion; only then does a need man­i­fest itself. This is as true for EDP as it is for any oth­er tech­nol­o­gy or imple­ment. The pre­ma­ture con­clu­sion that any imple­ment becomes nec­es­sary just because it mere­ly exists is a dan­ger­ous­ly nor­ma­tive one, name­ly the false con­clu­sion that the mere exis­tence of some­thing implies the neces­si­ty of its exis­tence, as philoso­phers would put it. Or, as an offi­cial of the Ger­man Min­istry, when it was still based in Bonn, for­mu­lat­ed it dur­ing his car­ni­val speech: Pota­toes must be eat­en sim­ply because they are there.
Let me rec­ti­fy this state­ment by say­ing: Only if we are hun­gry and pota­toes are an appro­pri­ate means of sat­is­fy­ing hunger will we have to eat them. If this is not the case, they won’t be eat­en, least of all will they be eat­en only on account of their being there.

This means that we need a nor­ma­tive jus­ti­fi­ca­tion if we want EDP to help us out of the emer­gency in which traf­fic mat­ters cur­rent­ly find them­selves. We could find such a jus­ti­fi­ca­tion in the goals of “achiev­ing more, more uni­for­mi­ty and improv­ing speed”. This would, how­ev­er, not be the most fun­da­men­tal jus­ti­fi­ca­tion pos­si­ble. We can also and this may sound sur­pris­ing to some appeal to jus­tice in order to show that the three aspects men­tioned above (enor­mi­ty, uni­for­mi­ty and speed) are involved with jus­tice, hence can be fun­da­men­tal­ly justified.

Postulates of Justice


As a start­ing point, let us dis­cuss enor­mi­ty. Jus­tice is there for every­one. If the quan­ti­ty of all those who expect jus­tice is large then jus­tice can be regard­ed as a bulk com­mod­i­ty, even if this may sound strange. Con­se­quent­ly, infor­ma­tion on statutes and case law is also at the dis­pos­al of every­one. For this rea­son the Asso­ci­a­tion for Com­put­ing in the Judi­cia­ry Con­gress’ mot­to some years ago was “Free Law for Free Cit­i­zens”, and last year it was “Free Jurispru­dence for Free Cit­i­zens”. Jus­tice is there­fore acces­si­ble for every­one, and if it were to become a lim­it­ed resource, we would find our­selves con­front­ed with a prob­lem of injustice.

By the way, the max­im “suum cuique tribuere” is evoca­tive of exact­ly the fol­low­ing idea: If every­body is giv­en what she/he is enti­tled to, then every­body has a right to justice.
And if every­body is enti­tled to this kind of jus­tice and the judi­cia­ry sys­tem acts accord­ing to the pos­tu­late which con­fers the right to jus­tice to every­one, we can say (and this may sound strange but is appro­pri­ate): The dis­course about jus­tice is a mat­ter of mass com­mu­ni­ca­tion. A tra­di­tion­al metaphor aris­es in this con­text. The Old Tes­ta­ment speaks of jus­tice as a stream that shall flow through the coun­try. We could adjust the term to our present sit­u­a­tion and say: like a flow of traffic.
All that means from a nor­ma­tive per­spec­tive is that if we can­not guar­an­tee the large-scale admin­is­tra­tion of jus­tice with­out EDP and there are var­i­ous rea­sons that jus­ti­fy its serv­ing role (I lay stress on the term serv­ing in this con­text) its appli­ca­tion is strong­ly advo­cat­ed on the basis of jus­tice. Thus the ques­tion “Shall we apply EDP?” los­es its mere­ly prac­ti­cal mean­ing of where and when are we sup­posed to install what kind of machines? Quite on the con­trary: We now encounter a fun­da­men­tal idea for which oppo­nents of EDP have yet to find an ade­quate alter­na­tive. We thus observe a momen­tous shift in the empha­sis of argu­men­ta­tion between oppo­nents and pro­po­nents of EDP. The pre­req­ui­site for this shift is always that sup­port­ers of EDP are con­cerned enough to make the attempt of find­ing a fun­da­men­tal justification.


The next point in our dis­cus­sion on traf­fic and traf­fic mat­ters is uni­for­mi­ty. Here we find a close con­nec­tion between jus­tice and uni­for­mi­ty. As we know, jus­tice in the tra­di­tion­al sense com­pris­es two aspects, one is mate­r­i­al jus­tice, and the oth­er is for­mal justice.
Mate­r­i­al jus­tice accord­ing to “suum cuique tribuere” is intend­ed to pro­vide every­one with the amount of jus­tice that they believe they are enti­tled to. For­mal jus­tice aims for the equal treat­ment of equal cas­es, a prin­ci­ple inher­ent in the con­sti­tu­tion­al law in Ger­many. This claim had already been recog­nised in ancient times. Cicero described it as fol­lows: Valeat aequitas quae in paribus cau­sis paria iura desider­at (the prin­ci­ple of aequitas should apply in order that equal legal con­se­quences for equal facts are ensured). From this under­stand­ing, uni­for­mi­ty is a prin­ci­ple of jus­tice. Con­se­quent­ly (and here we encounter a recur­rence of thought), if EDP as a mat­ter of prin­ci­ple turns out to be a sub­stan­tial con­tri­bu­tion to the equal treat­ment of cas­es, it will again prove to be a neces­si­ty in terms of jus­tice. Much speaks for such an approach. Equal treat­ment of equal cas­es requires, for instance, a com­pre­hen­sive overview of all the cas­es already adju­di­cat­ed. In view of the large quan­ti­ty of cas­es to be adju­di­cat­ed, it should be clear that only effec­tive infor­ma­tion tech­nol­o­gy would be able to tack­le this task, and thus even the cre­ation of data­bas­es is per­ti­nent to justice.
By the way, at an infor­mal meet­ing yes­ter­day it was recount­ed to me that a sim­i­lar idea had been one of the rea­sons behind the ini­ti­a­tion of the association’s sym­po­sium on traf­fic. The ini­tia­tors who were from Ham­burg were well aware of the dif­fi­cul­ty that courts, when decid­ing traf­fic cas­es, adopt incon­sis­tent approach­es accord­ing to their geo­graph­i­cal loca­tion. They intend­ed to rec­ti­fy this sit­u­a­tion by improv­ing the flow of infor­ma­tion and thus ensur­ing a uni­form appli­ca­tion of justice.


Jus­tice and Speed many will ask what the con­nec­tion between these terms could be. It is a clos­er one than we would at first think. The British legal sys­tem illus­trates this in a say­ing: Jus­tice delayed is jus­tice denied. Jus­tice which is admin­is­tered or grant­ed too late can be con­sid­ered a denial of jus­tice. The Euro­pean Court of Jus­tice for Human Rights has point­ed to this fact on sev­er­al occa­sions. This aware­ness is also con­spic­u­ous­ly expressed in the Latin phrase “bis dat qui cito dat” (he who gives quick­ly, gives dou­bly). The pas­sage of time deval­ues some­thing which has not been grant­ed on time. And as we don’t have an infi­nite amount of time at our dis­pos­al, jus­tice which is grant­ed too late may turn, in the worst case, into injus­tice. If we con­tin­ue to use traf­fic-relat­ed ter­mi­nol­o­gy we might say: Jus­tice is a just-in-time-con­cept (this is a dar­ing expres­sion, but it is cor­rect in the lit­er­al sense of the word). As a con­se­quence we can again say: If EDP is to make a worth­while con­tri­bu­tion to an accel­er­at­ed judi­ca­ture, its appli­ca­tion is com­pul­so­ry for rea­sons of justice.
It is a fact that EDP has the poten­tial to make a con­tri­bu­tion of this kind. How­ev­er, this is not always the case. Like the points dis­cussed pre­vi­ous­ly, EDP’s capa­bil­i­ty to reach these goals is only cor­rect in prin­ci­ple. EDP as such is not guar­an­teed to reach these goals. This rais­es a fun­da­men­tal ques­tion: At issue is not the use of EDP as such, but the effi­cient imple­men­ta­tion of EDP, of an EDP whose con­crete use demon­strates that it does in fact ful­fill the role it must (and is not mere­ly per­mit­ted to as many peo­ple would say) play for the above men­tioned nor­ma­tive reasons.

The Borderline between Efficient and Inefficient EDP

As men­tioned above, EDP is like­ly to keep its promise with regard to enor­mi­ty, uni­for­mi­ty and speed only if it is embed­ded in effi­cient data pro­cess­ing sys­tems. If this is not the case then pro­cess­ing back­logs may occur. The bulk of incom­ing cas­es can­not be ade­quate­ly coped with with­in the time required. The term “world wide wait” which is some­times used instead of “World Wide Web” reminds us of the fact that inef­fi­cien­cy is char­ac­ter­is­tic of the imple­men­ta­tion sce­nar­ios of non-scal­able sys­tems, (and this is not con­nect­ed to the stan­dards set for the WWW). Para­dox­i­cal­ly enough, court pro­ce­dures may even decel­er­ate on account of the imple­men­ta­tion of EDP. This is clear­ly exem­pli­fied by the first attempt under­tak­en some time ago to autom­a­tize an order of pay­ment procedure.
We also encounter inef­fi­cient infor­ma­tion sys­tems which ren­der find­ing rel­e­vant infor­ma­tion impos­si­ble. Thus a uni­for­mi­ty of treat­ment can­not be guar­an­teed due to these inef­fi­cient infor­ma­tion systems.

As a con­se­quence, we have to dif­fer­en­ti­ate (and here­in lies the main dis­tinc­tion) between bad­ly imple­ment­ed and effi­cient­ly imple­ment­ed EDP. It is not easy to know in advance how this dis­tinc­tion should be made. Nat­u­ral­ly, we are all the wis­er lat­er. The prin­ci­ple: “The test of the pud­ding is its eat­ing” is always reli­able. This is not, how­ev­er, a plan­ning strat­e­gy. The aim is rather to be able to dif­fer­en­ti­ate the one from the oth­er as ear­ly as pos­si­ble. And this is the basic plan­ning prob­lem which does not become appar­ent at all if we regard the imple­men­ta­tion of EDP as an unques­tion­able good.
There is cer­tain­ly no plan­ning sce­nario which could auto­mat­i­cal­ly enable us to detect inef­fi­cient EDP in the first place. Still, there are some promis­ing indi­ca­tors in this field. They can be found by ask­ing spe­cif­ic ques­tions and, if these can­not be answered by a strate­gic response, then the EDP sce­nario has to be han­dled with care.
The fol­low­ing ques­tions pro­vide an exam­ple of what this means in method­i­cal terms:

Are there any proven benefits to be gained from the new technique?

The ques­tion whether the rec­om­mend­ed sys­tem offers proven ben­e­fits com­pared to cur­rent oper­a­tional pro­ce­dures is one that must always be posed. For instance, when ask­ing this ques­tion with regard to the imple­men­ta­tion of elec­tron­ic files, we will soon per­ceive var­i­ous dis­ad­van­tages when we com­pare this method with that of tra­di­tion­al fil­ing. Intu­itive under­stand­ing of the “old” medi­um in com­par­i­son with the “new” one is an aspect of this dis­cus­sion. Some elec­tron­ic files may seem defi­cient when this approach is tak­en. These defi­cien­cies can be reme­died. If this is not done then we find our­selves from the very start con­front­ed with inef­fi­cient­ly imple­ment­ed technology.

Has the durability of the new technique been taken into consideration?

What about the manageability of the new technique?

Dig­i­tal sig­na­tures are illus­tra­tive exam­ples for ques­tions of man­age­abil­i­ty. It is indu­bitable that we need this new tech­nol­o­gy (this is in fact con­nect­ed with the elec­tron­ic file, since every entry and any change in it has to be authen­ti­cal­ly doc­u­ment­ed in the file). Even so, the imple­men­ta­tion of dig­i­tal sig­na­tures may cre­ate con­di­tions that ham­per the easy mas­tery of this basic tech­nol­o­gy. This is an indi­ca­tion of the fact that its present-day imple­men­ta­tion may not be the most effec­tive one. To con­clude that leg­is­la­tion is respon­si­ble for this sit­u­a­tion does not help mat­ters in the least.

Traffic as an EDP System

Let us reca­pit­u­late: EDP in prin­ci­ple con­forms not only to traf­fic cas­es but also to traf­fic. Let us do a cross-check: Could we ever imag­ine traf­fic as exist­ing with­out EDP? I don’t think so. With­out EDP, traf­fic could not func­tion prop­er­ly with regard to mass, uni­for­mi­ty and speed. I am well aware of per­haps repeat­ing com­mon­places, but some­times shared knowl­edge needs to be men­tioned in order to arrive at a sys­tem­at­ic frame­work. Let me give you a few key­words: Cars have mean­while evolved into “rolling infor­ma­tion sys­tems”. Net­worked infor­ma­tion sys­tems con­trol traf­fic. To put it in exag­ger­at­ed terms: Traf­fic has itself become an EDP system.

Chances and Risks

After hav­ing tried to arrive at some fun­da­men­tal insights, let us now take a look at the pos­si­bil­i­ties and risks of traf­fic-relat­ed EDP.
First of all we have to clar­i­fy the mean­ings of “pos­si­bil­i­ty” and “risk”. A risk looms large behind every­thing that seems to be a pos­si­bil­i­ty. Thus pos­si­bil­i­ties and risks are two sides of the coin. If we con­sid­er only the pos­si­bil­i­ties of EDP with­out keep­ing its risks in mind, fun­da­men­tal issues will be ignored (and vice ver­sa). Both aspects togeth­er form a com­pre­hen­sive pic­ture. Let me there­fore cite some cru­cial exam­ples for pos­si­bil­i­ties and risks.

The Network

Nowa­days the infor­ma­tion sys­tems that we encounter are net­works. This has become a fun­da­men­tal prin­ci­ple of these sys­tems. It is gen­er­al­ly regard­ed as an improve­ment, and right­ly so. Iso­lat­ed infor­ma­tion islands no longer exist. Data exchange between het­ero­ge­neous appli­ca­tions via the world wide net­works has become com­mon prac­tice instead. The inter­net and the vision of a “seman­tic web”, a net­work that has been giv­en a well-defined mean­ing which facil­i­tates infor­ma­tion exchange, has been very explic­it­ly demon­strat­ed by the W3C-con­sor­tium. This was the brain­child of Tim Bern­ers-Lee, who, along with Robert Cail­lau, was one of the two “fathers” of the World Wide Web. Tim Bern­ers-Lee is con­vinced that this tech­nol­o­gy does not have any inher­ent mean­ing and so requires mean­ing to be fed into it. Sense is con­nect­ed to mean­ing, and this mean­ing needs to be anchored to the World Wide Web. At present, this is one of the most excit­ing issues in legal infor­mat­ics. The Ger­man world of jurispru­dence has not yet ade­quate­ly real­ized the impor­tance of this issue, and con­se­quent­ly our field of research still lies out­side the sci­en­tif­ic main­stream. If these key­words should appear to be too cryp­tic, help can be obtained from the W3 web­sites which can be con­sult­ed as one would a large library.

The Cooperative Connection

Nowa­days, indi­vid­ual EDP sys­tems are part of a coop­er­a­tive coher­ent sys­tem (and can only be ade­quate­ly under­stood as such). This fact is a chance and is also a cen­tral risk. It is essen­tial to find a frame­work where the risks and the pos­si­bil­i­ties are prop­er­ly bal­anced. This pos­tu­late refus­es an “either / or” log­ic. In con­trast to this, we often encounter EDP dis­cus­sions that are, strict­ly speak­ing, struc­tured accord­ing to this log­ic. This strat­e­gy is method­i­cal­ly mis­lead­ing . My lec­ture there­fore intend­ed to show that EDP in traf­fic mat­ters is essen­tial, not for its own ben­e­fit and not as an arbi­trar­i­ly imple­ment­ed device, and that it is imper­a­tive to find a bal­ance between diver­gent expec­ta­tions and objec­tives. It would be a great achieve­ment if this prin­ci­ple could under­lay any dis­cus­sion about the use of EDP in legal matters.

The Language of Images

Anoth­er fur­ther mix­ture of pos­si­bil­i­ties and risks is pre­sent­ed in the way we han­dle images. At the entrance of this hall we encoun­tered an inter­est­ing tech­nol­o­gy which serves to visu­al­ly recon­struct road acci­dents. When look­ing at these images, we are remind­ed of dialec­tics in the clas­si­cal sense of the word: On the one hand, we are encoun­ter­ing a use­ful tech­nol­o­gy, a pos­si­bil­i­ty. On the oth­er hand, there are also risks lurk­ing here. Images, as we all know, exert a pow­er­ful influ­ence on us. Since we tend to suc­cumb to this prin­ci­ple, we live in the age of visu­al­iza­tion. In Amer­i­can traf­fic courts, the recon­struc­tion of road acci­dents is shown in film form. The illu­sion that is there­by cre­at­ed is that real­i­ty has been cap­tured and brought into the court room, an illu­sion which is clear­ly not true. Images are only reflec­tions of real­i­ty, not real­i­ty itself. The images fol­low their own sug­ges­tive rhetoric. Prof. Kroe­ber-Riel, my col­league from Saar­land Uni­ver­si­ty, is the author of a work of ref­er­ence enti­tled “Visuelle Kom­mu­nika­tion” (Visu­al Com­mu­ni­ca­tion). If we cease to be aware of the dif­fer­ence between real­i­ty and images, and if we don’t reflect on it in a nuanced way, then a dan­ger­ous appli­ca­tion of tech­nol­o­gy with unpre­dictable con­se­quences may occur. As this trend towards visu­al­iza­tion does not yet play a pre­dom­i­nant role in legal pro­ce­dures in Ger­many, there is still enough time left for us to pre­pare for it through prepara­to­ry reflec­tion. This might be a com­mon issue for the association’s sym­po­sium on traf­fic and the Asso­ci­a­tion of Com­put­ing in the Judi­cia­ry: The lan­guage of images in traf­fic matters.

Blind Trust

The last kind of risk I would like to dis­cuss is a fun­da­men­tal one. We could call it “the risk of blind trust”. This atti­tude is espe­cial­ly dan­ger­ous as it tends to be nur­tured by the appar­ent­ly smooth func­tion­ing of EDP. The risk that we would nev­er trust soft­ware which does not func­tion is min­i­mal. Yet we auto­mat­i­cal­ly believe that all is well when EDP appar­ent­ly works smooth­ly. But this is not nec­es­sar­i­ly true. As a con­se­quence, a field of study with regard to “trust­ed com­put­ing” has evolved in infor­mat­ics. Its cen­tral ques­tion is: “What makes us trust appar­ent­ly func­tion­ing EDP?” It would be advis­able to make this top­ic, which is spe­cif­ic to infor­mat­ics, rel­e­vant for our con­cerns as well. Let me give you an exam­ple of where the prob­lem lies: Some time ago incor­rect­ly pro­grammed pock­et cal­cu­la­tors were giv­en to pupils in the U.S. in order to study the pupils’ reac­tions. It was only after a 25% devi­a­tion from the expect­ed result became vis­i­ble that the pupils point­ed out the cal­cu­la­tion error, and this was done only very hes­i­tant­ly. This behav­ior not only occurs in arti­fi­cial­ly induced sit­u­a­tions, it also aris­es in real life. If you take a cal­cu­la­tor and extract the root of a num­ber, say 3, and then square the result, you should obtain the same num­ber you start­ed off with, but often enough this is not the case. This is an indi­ca­tion of the calculator’s inner arith­metic being dif­fer­ent from what strict­ly for­mal math­e­mat­ics require. The dis­turb­ing ques­tion as to which result we can trust remains.

Is the machine taking over?

What will the final result be in this sit­u­a­tion? The pos­si­ble con­se­quences may be indeed seri­ous. We may lose our own pow­ers of judg­ment. The machine will take over. I fear this has already hap­pened in many areas of our every­day life. If we were to take away somebody’s EDP-machine, would they still be able to cope with the task they’ve been set? A sci­ence fic­tion author has tak­en this issue as the mate­r­i­al for an impres­sive sto­ry: Some­time in a future mil­len­ni­um a giant space­ship los­es its ori­en­ta­tion in the galaxy. Its com­put­er has bro­ken down. It is not pos­si­ble to cal­cu­late the para­me­ters of its return course. Thou­sands of peo­ple on board there­fore pre­pare for their immi­nent death until some­body says that for­mer­ly peo­ple were able to cal­cu­late cours­es with­out the help of machines. The crew now orga­nizes itself into sev­er­al units, each of which is assigned the task of solv­ing a part of the arith­meti­cal prob­lem. Each result is then passed on to the next unit, and in the end the groups man­age to work out the return course. But what, asks the author, if no one had been able to cal­cu­late any more? There­in lies the real dan­ger. And in con­sid­er­ing this, we are approach­ing the final key point in our analy­sis: The limits.

The Limits

We should always con­sid­er that there are lim­its to every instru­ment which have to be assessed. But besides these more com­mon lim­its there are oth­er lim­its in the case of cer­tain tools (com­put­ers being per­haps the most pro­to­typ­i­cal of them) which cause mankind to re-assert itself. In our con­text this is the case when the ward (EDP) wants to become the guardian, or when EDP refus­es to obey com­mands any longer. The prin­ci­ple of human self-asser­tion in the face of the threat posed by machines when used judi­cious­ly can be very help­ful in assess­ing EDP projects. For exam­ple, a state­ment such as “the EDP won’t per­mit it” ought to imme­di­ate­ly trig­ger reflec­tion as to whether the machine is sig­nal­ing its wish to take over com­mand. We should devel­op a very high degree of sen­si­tiv­i­ty to this kind of sit­u­a­tion and go to the trou­ble of pre­vent­ing the essen­tial­ly use­ful machine on which we actu­al­ly depend from dom­i­nat­ing us.