Social Policy and Cognitive Enhancement: Lessons from Chess · example of chess for the regulation of cognitive en-hancement more generally in education and the pro- ... performance

ORIGINAL PAPER

Social Policy and Cognitive Enhancement: Lessonsfrom Chess

Emilian Mihailov & Julian Savulescu

Received: 17 October 2017 /Accepted: 1 February 2018 /Published online: 22 February 2018# The Author(s) 2018. This article is an open access publication

Abstract Should the development of pharmacologi-cal cognitive enhancers raise worries about doping incognitively demanding activities? In this paper, weargue against using current evidence relating to en-hancement to justify a ban on cognitive enhancersusing the example of chess. It is a mistake to assumethat enhanced cognitive functioning on psychometrictesting is transferable to chess performance becausecognitive expertise is highly complex and in largepart not merely a function of the sum specific sub-processes. A deeper reason to doubt that pharmaco-logical cognitive enhancers would be as significant inmind sports is the misleading parallel with physicalenhancement. We will make the case that cognitiveperformance is less mechanical in nature than phys-ical performance. We draw lessons from this caseexample of chess for the regulation of cognitive en-hancement more generally in education and the pro-fessions. Premature regulation runs the risk of creat-ing a detrimental culture of suspicion that ascribesunwarranted blame.

Keywords Cognitive enhancement . Mind sports .

Chess . Cognitive complexity . Cognitive expertise .

Policy . Trustworthiness

Introduction

At the 2008 Chess Olympiad in Dresden, officials re-quested that Vassily Ivanchuk submit to a drug test afterlosing to Gata Kamsky. The media reported that Bhestormed out, kicked a pillar in the lobby, pounded acountertop in the cafeteria with his fists and then vanishedinto the coatroom.^1 Perhaps this was adding insult toinjury, angering an exhausted and volatile grandmaster.But since Ivanchuk refused to take a urine test, theofficials treated his refusal as a confirmation of doping.

The decision caused outrage among chess profes-sionals. Many elite players believe that doping accusa-tions are an insult to their profession and argue thatperformance enhancing substances bring no benefits tothe sport. In an open letter of support of Ivanchuk,grandmaster Alexei Shirov accused the World ChessFederation (Federation Internationale des Echecs, knownas FIDE from its French acronym) of unjustifiably im-posing anti-doping rules just to meet the requirements ofthe International Olympic Committee for labelling chessas an Olympic sport.2 In response, the FIDE President

Neuroethics (2018) 11:115–127https://doi.org/10.1007/s12152-018-9354-y

1 http://www.spiegel.de/international/zeitgeist/outrage-over-ivanchuk-the-great-chess-doping-scandal-a-595819.html2 See the press release here: http://en.chessbase.com/post/alexei-shirov-let-us-ban-fide-

E. Mihailov (*)Research Centre in Applied Ethics, Faculty of Philosophy,University of Bucharest, Splaiul Independentei no. 204,060024 Bucharest, Romaniae-mail: [email protected]

J. SavulescuOxford Uehiro Centre for Practical Ethics, University of Oxford,Suite 8, Littlegate House, St Ebbes Street, Oxford OX1 1PT, UKe-mail: [email protected]

Kirsan Ilyumzhinov argued that anti-doping rules werenecessary to protect chess against cheating:

BChess as a sport itself deserves the clean compe-tition of the players, devoid of falsifications,cheating and doping. From the very beginning,we were well aware of that in chess we wouldnot have to cope with steroids or other hormonesused in the physical sports, but at the same timethe scientific research identified several sub-stances that could affect the mental performanceof a chess player.^3

The outrage of elite chess players may be motivatedby professional integrity,4 but there is a tension betweenthe scepticism of professional players and the belief ofFIDE that pharmacological cognitive enhancementcould offer advantages. Scepticism about enhancementmay be driven by assumptions equating doping to phys-ical, mood, and motivational enhancement.5 But thismisses the germane challenge of pharmacological de-velopments which are shown to have beneficial effectsdirectly on basic cognitive functions.

Due to the expansion of prospective technologicaland pharmaceutical enhancements, the use of perfor-mance enhancing substances in sports is fiercely debat-ed. However, much of the debate has focused on phys-ical sports, while discussions about cognitive enhance-ment have focussed on risk, benefit and the publicinterest, particularly in relation to education and em-ployment [2, 3]. Now, there is a growing tendency totake the development of enhancement substances asproviding means of mental doping in cognitive sports.The International Mind Sports Association, which reg-ulates bridge, draughts, go, xiangqi, also adheres to theWorld Anti-Doping Agency’s policy on prohibited sub-stances. Should the development of smart drugs for thegeneral population outside therapeutic contexts raise

worries about doping in cognitively demanding sportsand games? More generally, should cognitive enhance-ment be banned in complex professions or education?

There are reasons to resist this tendency, at least forthe present time. In this paper, we use the test case ofchess and we argue that a ban is based on a misunder-standing of the evidence about enhancement, the generalconnections between cognitive processes and the com-plexity of mental performance. Cognitive enhancementsubstances should, nevertheless, be monitored, just ascaffeine and codeine are and further evidence gathered.Ethical analysis will be necessary to evaluate this evi-dence and decide whether cognitive enhancement ismorally permissible, but in the first instance we arguethat it is a mistake to jump to conclusions about phar-macological enhancement of complex cognitive perfor-mance, as in the case of chess, from evidence of modestimprovements outside that activity or in lay people. Wewill argue that there is a stark discrepancy between thehigh expectations and the actual enhancement effects.The realistic picture we outline on the effectiveness andmodus operandi of pharmacological cognitive enhance-ment will help assess whether there is direct and indirectevidence for an impact on performance in chess. Wefurther argue that we should avoid assuming that en-hanced cognitive functioning on psychometric testing istransferable to chess performance because cognitiveexpertise is highly complex and in large part not merelya function of the sum specific sub-processes. A deeperreason to doubt that pharmacological cognitive en-hancers would be as significant in mind sports is themisleading parallel with physical enhancement. We willmake the case that cognitive performance is less me-chanical in nature than physical performance.

If we are right, the case of chess provides importantlessons to be learned for the broader social context ofusing cognitive enhancement. We cannot form policy oncomplex questions about cognitive performance in com-petitive activities, education, or the workplace from sim-ple lab tests that have questionable ecological validity.Premature regulation runs the risk of creating a detrimen-tal culture of suspicion that ascribes unwarranted blame.

Belief in Enhancement

According to the ChessWADAAnti-Doping Policy, themost relevant banned substances for chess are amphet-amine derivatives (Adderall, Ritalin), ephedrine and

3 http://gambit.blogs.nytimes.com/2008/12/26/world-chess-federation-president-addresses-controversies/?_r=04 According to the WADA [1] Anti-Doping Rule Violations (ADRVs)Report, there were no anti-doping rule violations in chess.5 After the Ivanchuk scandal, many top level grandmasters were askedby the media what chess doping means to them. Here are some typicalanswers: BI think, chess is more intellectual than physical kind of sport.There is no point in speaking about chess doping.^ (Peter Leko) BIthink doping doesn’t assist chess-players that much and I doubt ifanybody uses it at the top-level tournaments. Chess isn’t sport in itsoriginal scope, it’s not an Olympic kind of sport like track-and-field orgymnastics.^ (Shakhriyar Mamedyarov) See: https://chessdailynews.com/whats-chess-doping-to-you/

116 E. Mihailov, J. Savulescu

methylephedrine, pseudoephedrine, and Modafinil. No-tably, caffeine and codeine are not prohibited, but figurein the Monitoring Program.6 The reasoning is probablythat a normal dosage of caffeine or codeine is highlyunlikely to present significant benefits to players. TheFIDE believes that cognitive enhancers have the poten-tial to be of benefit in chess, and has thus implicatedModafinil, Adderall and Ritalin. In therapeutic contexts,Modafinil is used to treat narcolepsy, shift work sleepdisorder and excessive daytime sleepiness, by improv-ing wakefulness. Ritalin and Adderall are nervous sys-tem stimulants used for ADHD treatment, a disordercharacterised by concentration and difficulty with im-pulse control.

Many studies have shown that Modafinil, Adderalland Ritalin benefit people in non-therapeutic contexts.Modafinil enhances performance on tests of digit span,visual pattern recognition memory, spatial planningand stop signal reaction time [4, 5]. Additionally, thereis evidence that non-sleep-deprived individuals givenModafinil have higher accuracy and marginally fasterresponse times on tests of conceptual ability [6]. Ad-ministration of methylphenidate (Ritalin) has beenshown to improve problem solving abilities, spatialworking memory, declarative memory consolidationand the ability to divide and relocate attention [7–9].Equally important, stimulants and Modafinil have sig-nificant non-cognitive effects on motivation, moodand alertness, factors which are closely related tooverall cognitive performance [10].While many pro-fessional players may not be aware of these kinds ofeffects, FIDE has used such evidence as a basis to bancognitive enhancers.7

The evidence used by FIDE refers to enhancementeffects outside mind sports, but a recent study docu-ments for the first time enhanced performance in chessby methylphenidate and Modafinil. In randomizedcrossover trial, Andreas Franke and his colleagues [11]administered Modafinil, methylphenidate and caffeineor placebo to participants who played against chesssoftware. Results showed an increased performance inaverage scores for all three substances, implying that theplayers performed 3–4 percentage points above placebo

performance. We will discuss below how relevant thisevidence is and whether it is sufficient to justify a ban.

Another potential cognitive enhancer in chess istranscranial direct current stimulation (tDCS). TDCSstimulates the brain by passing a low intensity currentthrough two electrodes placed over the head whichmodulates neuronal activity. A recent meta-analysis re-vealed a small but significant effect in healthy partici-pants of left dorsolateral prefrontal cortex stimulationcoupled with working memory training [12]. Such de-vices are widely used in computer games [10]. They arenot prohibited.

Effectiveness of Pharmacological CognitiveEnhancement

The media uses metaphors like Bmiracle drug^ andBbrain steroid^ to describe cognitive enhancers [13].These contribute to an exaggerated image of the effec-tiveness of enhancers. Part of the debate has mislead-ingly focussed on extreme forms of bioenhancementthat promise radical transformations of human capaci-ties, especially drugs that might dramatically increaseintelligence to super-human levels, the prospect of bi-onic athletes, and biotechnological interventions thatwould extend our life span indefinitely [14–16]. Inresponse, some have argued that there is a need torefocus on normal range human enhancement [17]. Aswe will argue, there is a stark discrepancy between thehigh expectations and the actual effects.

Overall, no sweeping general conclusion can bedrawn regarding the effectiveness of pharmacologicalcognitive enhancement even in the laboratory setting.There is no general effect exhibited by smart drugs, aseach drug in its own right and dosage will have differenteffects [18]. In their meta-analysis of randomized con-trolled trials of Methylphenidate, Repantis et al. [19]found some effect on memory, but no significant im-provements in attention or executive function. In otherexperiments, Methylphenidate did not affect spatialworking memory or planning [8]. Farah et al. [20]confirm that Methylphenidate and amphetamine im-prove memory when the retention interval betweenstudy and test was longer than an hour, but the evidenceon executive functions was inconclusive. Fond et al.[21] identified enhancement effects of Methylphenidateon working memory, but again no significant effect onattention and executive functions.

6 TheWADAProhibited List andMonitoring Program can be found at:http://list.wada-ama.org/7 See the FIDE Chess WADA – Anti-Doping Policy, Nutrition andHealth at http://www.fide.com/FIDE/handbook/WADA%20Anti%20Doping.pdf

Social Policy and Cognitive Enhancement: Lessons from Chess 117

With respect to Modafinil, the reviews of Fond et al.[21] and Repantis [22] found moderate positive effecton attention, but none for memory. Other reviews con-clude that although there is clear evidence of enhancedexecutive function and memory in sleep-deprived indi-viduals, for well rested individuals there is a large num-ber of null results and occasionally even cognitive im-pairment [20]. Battleday and Brem [23] also highlightthat most studies did not report any effect of Modafinilintake on problem-solving abilities, and a few evenreported impairments in divergent creative thinking. Itseems that Modafinil preferentially affects attention andexecutive function of lay population, with no significantinfluences on memory and problem-solving abilities.

It is also important to highlight that enhancementeffects are not always upgrades over normal perfor-mance. Studies which measure cognitive performancedo not employ the same enhancement model, raisingmethodological problems in extrapolating from over-views. We should distinguish between the ‘better thanbaseline’ model, used to measure improvements in nor-mal functioning, and ‘sustaining the baseline’ model,used to measure the extent to which normal functioningcan be preserved in adverse conditions. Modafinil wasmainly tested in studies with sleep-deprived individuals[22], that is, sustaining the baseline.

The high expectations of the potential of pharmaco-logical cognitive enhancement fuelled by the media andpopular culture8 should be tempered. Much of the cog-nitive enhancement scientific literature is 15 to 20 yearsold and that more recent investigations are often equiv-ocal. Current evidence suggests only modest effects onsome cognitive sub-functions measured by fairly simplespecific laboratory tasks, with no effects on other cog-nitive sub-functions and even impairments. Farah et al.[20] refer to this pattern of limited improvements onsome specific tasks and impairment on others as beingBfamiliar^ for pharmacological cognitive enhancers.

Enhancement effects are modest in healthy individ-uals, but even this size can be misleading because itdiscards the variations within subgroups. If we look intosubgroups whose baseline performance is poorest theeffects becomemore significant. In many studies, Meth-ylphenidate improved working memory in people whowere low performers, though on the other hand, it im-paired performance in participants with high working

memory baseline [24, 25]. There is also evidence thatModafinil is less effective in high-performing individ-uals than in low-performing individuals [26], as well asin high-IQ versus low-IQ individuals [4]. These resultssuggest another Bfamiliar^ pattern for pharmacologicalcognitive enhancement –more consistent improvementsin low performing individuals and no significant influ-ence or even impairment on individuals with a highperformance baseline. It is, importantly, possible thatputative cognitive enhancers impair certain functionsin elite chess players.

To get an estimate about the size of some pharmaco-logical enhancement effects, naps are more effective inmaintaining performance than Modafinil and amphet-amine during periods of sleep deprivation in healthyindividuals [27]. The more realistic picture we outlinedabout the effectiveness and modus operandi of pharma-cological cognitive enhancement helps to draw morereliable implications for professional chess performance.

Implications for Chess Performance

Direct Evidence?

We have argued that the existing evidence of enhance-ment effects on cognitive functions is modest in general,and there is the possibility of impairment in certaindomains. Moreover, there is no direct evidence relatingto elite professional chess performance. Almost all rel-evant evidence is taken from studies on healthy laysubjects performing basic psychometric tests.

There is one recent study showing enhancement ef-fects on chess related tasks, but subjects were in factamateurs, which questions its ecological validity forprofessional performance. What is striking in this studyis how the enhancement effect was supposedly mediat-ed. It was not improved memory, accuracy or planning,but what Franke and his colleagues described as Bmorereflective decision making^ ([11], p. 257). The averagereflection time per game was considerably higher com-pared to placebo during middle game (moves 11 to 25).It seems plausible that Bmore reflective decisionmaking^ did not make players Bdeeper^ thinkers, asmedia outlets claimed.9 Rather, it seems that the effect

8 The film Limitless portrays an exaggerated version of the influenceof Modafinil.

9 See https://qz.com/905546/smart-drugs-like-modafinil-and-ritalin-can-be-used-as-doping-agents-to-enhance-performance-in-chess-games/

118 E. Mihailov, J. Savulescu

was to delay making a move. Subjects just took moretime for game analysis, presumably having the subse-quent effect of seeing more optimal moves and counter-moves, while players in the placebo condition movedfaster, without thoroughly assessing the positions. Im-provements were not direct outputs of enhanced thoughtprocess, but rather side-effects of a better impulse con-trol of the tendency to react quickly to the opponent’smoves.10 It is standard training to teach beginners totake their time before each move, as they move quicklywithout thinking too much, but this is unlikely to helpelite professionals who have in depth experience inallocating time to evaluate complex positions. Suchprofessionals can literally spend hours deliberating overposition and possible moves.

For these reasons, conclusions about pharmacologi-cal cognitive enhancement as a form of performanceenhancement in cognitive sports should not follow di-rectly from an analogy with physical performance en-hancement. In the case of essentially physical sports, theevidence points to strong enhancement effects in direct-ly relevant activity and is taken from studies with eliteand sub-elite athletes performing their specific sport (seefor example [28]). Moreover, it is a mistake to drawconclusions about complex cognitive activities likechess directly from mild cognitive enhancement effectsin a diverse population of lay people on simple psycho-metric tasks.

To advance the debate on mental performance, majorresearch needs to start with the influence of cognitiveenhancers on elite and sub-elite chess players performingspecific chess tasks. We should aim for a minimal eco-logical validity. In chess there is a clear system of ratinghow players excel, heavy opening theory, well organizedchess puzzles based on difficulty levels, and clear stan-dards of strategic thinking. Tests for measuring perfor-mance at chess specific tasks could be devised withregards to long term memory, pattern recognition andstrategic thinking, taking into account, as well, thestrength of players.

Even under strict laboratory conditions, enhancementeffects are generally minor and often the same as withconventional means. The single existing chess studyshowed caffeine to be just as effective as Modafiniland was slightly better than methylphenidate [11]. So

even if we get the same results for professional chessplayers as it is documented for lay people on tests ofopening moves, we should expect minor effects as well,which could more easily fade away outside experimen-tal conditions. Moreover cognitive enhancers like meth-ylphenidate and Modafinil appear to offer little advan-tage over caffeine, which is already legal.

Indirect Evidence?

One objection to our argument that we cannot infersignificant benefits to chess from current evidence isthat pointing to lack of direct evidence does not ruleout the possibility that cognitive enhancement is proba-bly achievable in an indirect manner [29]. Pharmaco-logical substances improve cognitive sub-functionssuch as attention, alertness, memory and informationprocessing. These cognitive sub-functions are used tosupport general cognitive abilities. These more integrat-ed cognitive processes determine the way in whichcomplex tasks of chess playing are performed. Forexample, pharmacological enhancement improvesmemory, a cognitive function which can improve thelearning process. Then, improvements in learning willpositively support the task of solving more chess puz-zles. The argument assumes that improvements at thelevel of cognitive sub-functions will translate to a higherlevel of cognitive functioning and, most decisively, theimprovements of the latter will translate to domainspecific and highly demanding cognitive processes.

While this objection may have some plausibility, it ispossible to argue equally in the opposite direction. Ascomplexity grows, it is plausible to assume that im-provements on basic memory tasks will taper off, afterwhich even impairments can be seen. After all, sincememory improvements are minor, why shouldn’t weassume that enhancement effects will phase out as wework using more integrated and complex cognitive pro-cesses? Maybe improvements are seen precisely be-cause task performance does not require exceptionaland integrated cognitive processing. By using an excep-tionally complex visual attention task, Rogers et al. [30]concluded that methylphenidate actually disrupts theallocation of attention between relevant and irrelevantfeatures of the environment. So we might indirectlyinfer that cognitive enhancers might impair performancein chess. Another study using a complex video game,which requires problem-solving strategies, found thatRitalin (methylphenidate) disrupted performance [31].

10 The study has an important limitation in this regard. It did notassessed participants for ADHD symptoms such as inattention, hyper-activity or impulsivity during the study.

Social Policy and Cognitive Enhancement: Lessons from Chess 119

In short, results of putative enhancements on integratedcognitive processes are generally negative [32].

Another version of this objection is that, while thereis no general effect of pharmacological enhancement oncomplex processes, we should distinguish different sub-stances. Stimulants (such as Ritalin and Adderall) seemto preferentially target memory with no significant ef-fects on other cognitive functions, while Modafinilseems to preferentially target attention and executivefunction with little or no significant effect on memory.This implies that no sweeping claim can be made aboutindirect enhancement. At best it can be argued that aparticular substance can preferentially target a particu-lar cognitive function. Some tasks may rely more onmemory, while others may be based more on executivefunction. We should not assume general correlationsbetween all pharmacological cognitive enhancers andall mind sports.

Moreover, the relation between a particular sub-stance and a particular cognitive function should bedetermined more precisely. It is not enough to argue infavour of enhancement effects from the premises that aparticular substance improves memory and that bettermemory improves chess playing. Memory and execu-tive function have many sub-domains, which in psycho-metric tests are measured independently from each oth-er. A pharmacological enhancer does not improve mem-ory in general, but specific sub-functions and not allmemory sub-functions improve performance at chess.Suppose that a particular substance has significant en-hancement effects on visual memory. There is in factevidence that visual memory is relatively unimportant tochess skill ([33]), whereas storage in long term memoryis central [34].11

However, a Bfamiliar^ pattern of pharmacologicalenhancement that we have highlighted is characterisedby more consistent improvements in low performingindividuals and no significant influence on individualswith high performance baseline or even impairment.Chess is a highly demanding mental activity and, notsurprisingly, its practice can be a useful tool to enhanceour intellectual abilities. In light of this, there are corre-lations between chess skill and IQ [36, 37]. Also, eliteplayers possess very strong long-term memory abilities.Chase and Simon [34] showed that the differences inability of chess players to recall positions are explained

by the experts’ storage of thousands of chunks of infor-mation in their long-term memories. Others estimatethat one needs to memorize about 100,000 openingmoves in order to reach high levels of expertise in chess[38]. It is unlikely that we will find professional chessplayers in the low baseline group, where the benefits ofpharmacological enhancement are to be found. More-over, there is no evidence of any cognitive enhanceraffecting long term memory at this level.

The other Bfamiliar^ pattern of pharmacological en-hancement is exhibited by limited improvements onsome specific tasks and impairment on others, callinginto question the status of the overall enhancement evenof specific enhancers. This undermines the idea thatimprovements of a particular function will bring benefitsto more highly integrated cognitive processes. Enhance-ment effects are also coupled with impairments of otherparticular functions, which may reduce the beneficialeffects or even de-enhance cognitive performance. Afew studies reported impairment of problem-solvingabilities after taking Modafinil. And most importantly,although Modafinil and methylphenidate increased re-flection time per chess game in the only ecologicallyvalid study of chess performance, resulting in morewins, it also led to more losses due to running out oftime as compared to placebo [11].

Here is one example of a potential pharmacologicalenhancer which we did not mention in our analysis, butnevertheless, it is famously related to chess performanceand nicely illustrates the point. The use of beta-blockers isbanned in archery, billiards, snooker. Beta-blockers caninhibit unwanted influences of arousal, anxiousness, orstress, by blocking adrenaline from activating a Bfight-or-flight^ response. It is shown that low rated and sub-elitechess players who exhibit heart rates in excess of 200/minand large increases in catecholamines tend to make moresimple mistakes [39]. The negative effects of adrenalineon the quality of play can be diminished by reducing theplayers’ heart rates and anxiety. On the other hand, onefamous self-experiment conducted by Helmut Pfleger, aGrandmaster and medical doctor, tells us the impairmentside of the story. Pfleger tested the effects of beta-blockerson himself in 1979, in a match with former world chesschampion Boris Spassky, by taking a tablet of Belocbefore the game. His heart rate went down to 40 to 65(in comparison with his normal range of 80–110) and helost almost immediately to Boris Spassky, after just twen-ty moves. Pfleger stated that his game collapsed when hisblood pressure plunged, drawing the harsh conclusion

11 Similar findings show that other cognitive sports, such as bridge, aresupported by working memory, but not by semantic memory [35].

120 E. Mihailov, J. Savulescu

that both mentally stimulating and mentally calmingmedication have too many negative effects [40].

Even though beta-blockers can enhance performanceof low rated and sub-elite players when a massive re-lease of adrenaline chokes the functioning of cognitiveskills, it is not clear whether elite chess players canbenefit. At least one elite player on one occasion didnot. Beta blockers may reduce the probability of makingsimple mistakes which are facilitated by anxiety. How-ever, when it comes to fine-tuned abilities of eliteplayers, the prospect of fatal impairment is present.Mentally calming medication can decisively numb eliteplayers’ ability to foresee potentially dangerous patternsand react accordingly. This simple beta blockerBexperiment^ is of course not a randomised controlledtrial but it is does suggest the potential for enhancers tobe damaging to elite performance.

General and Specialized Cognitive Abilities

The mind is not based on a delicate interdependence ofcognitive functions, such that one intervention will alterBsomething throughout all parts of the immeasurablewhole^, to use the expression of German philosopher,Johann Gottlieb Fichte.12 Cerebral hemispheres worktogether to maintain mental unity, as shown by split-brain studies [42], but the mind also contains a con-glomeration of independent systems. Announcementsof enhancements effects on memory, learning or reason-ing are based on imprecise general labels of cognitivefunctions. Undifferentiated talk about cognitive en-hancement is misleading without discussing which par-ticular cognitive function is investigated. For example,memory is not a monolithic brain system, as one has totake into account the distinctive neurobiological andneurochemical systems that make up, at the very least,semi-dissociable systems of memory [43], such as longand short termmemory. This point will help to highlightthat cognitive complexity also consists in developingand sustaining specialized cognitive abilities, which arerelatively independent from general thought processes.

Professional chess players are usually perceived ashaving superior intelligence, abstract thinking andmem-ory abilities. This might be true to some degree, but it isa misleading picture. Between experts and non-experts

there aren’t significant differences in their generalthought processes [44]. Master level players consideralmost the same number of possible moves, searchheuristics or engage in the same depth of search asweaker players. When required to recall positions withchess pieces that are randomly placed on the board,expert players have the same short-term memory asnon-experts.

What then makes the difference if not the grosscharacteristics of thought and memory processes? Thereis a vast body of data which explains cognitive expertiseby acquiring a large, complex, and flexible amount ofknowledge during domain-specific practice and experi-ence ([34], [45–47]). With extensive practice a special-ized content determined by a set of rules is stored in longterm memory. Experts are then able to identify, selectand combine chunks of information that are relevant fortheir cognitive tasks. For example, expert chess playersdo not engage in more intense cognitive processing thannon-expert players, but they are far better at identifyingspecialized patterns. In comparison, non-experts con-sume cognitive and time resources analysing irrelevantcontent [34].

In general, non-experts do a poor job at discardingirrelevant information, whereas experts master the abil-ity to process the meaningful information for a specificdomain. However, the playing field is levelled whencognitive tasks involve information and patterns outsidethe specific domain. As noted earlier, chess players donot perform better than non-experts at recalling non-specialized information. Counterintuitively, nor do su-per-memorizers. In one study, Ramon et al. [48] had theopportunity to test two subjects with exceptional successin the World Memory Championships. As expected,super-memorizers easily out-performed control subjectsat face-name learning tasks similar to those used ininternational memory competitions. However, theplaying field was levelled when all subjects performeda task outside the specialized domain. Despite theirsuperior performance at specialized tasks, super-memorizers did not differ from control subjects in non-specialized tests, such as face inversion recognition.

It seems that general cognitive abilities are consid-ered for the most part to be unrelated to expertise.Against the popular picture, Gobet et al. note that thereis not Ba single study that has shown that more skilledchess players outperform less skilled chess players onany psychometric test.^ ([33], p. 305) Others point outthat general intelligence is Beither unrelated or weakly

12 In his The Vocation of Man, Fichte says that one Bcould not remove asingle grain of sand from its place, without thereby (…) altering some-thing throughout all parts of the immeasurable whole^. ([41], p. 26)

Social Policy and Cognitive Enhancement: Lessons from Chess 121

related to performance among experts (…); factorsreflecting motivation (…) are much better predictors ofimprovement^ ([45], p. 280).13 One should avoid as-suming that enhanced cognitive functioning at psycho-metric tests is simply transferable to performance ofcognitive expertise.

The Mechanical Model of Enhancement

We pointed out earlier that judging pharmacologicalcognitive enhancement as a form of performance en-hancement in cognitive sports should not follow theexample of physical enhancement substances becausethe levels of relevant evidence are not even close. How-ever, there is a deeper reason as to why cognitive per-formance may not parallel physical performance. Phys-ical performance is more mechanical in nature thancognitive performance. Enhancement effects on physi-cal performance are more quantifiable, robust and sig-nificant, because the causal links of improvements aremore straightforward between enhancement interven-tions and single physiological mechanisms of directrelevance to performance.

The mechanical model holds that if we improve aspecific mechanism, or component of performance, wewill likely improve overall performance. This model iswell illustrated in endurance sports. Endurance athletesbenefit from increased haemoglobin levels, which in-crease their oxygen carrying capacity. Such improvedoxygen delivery improves athletes’ aerobic capacity,defined as the maximum amount of oxygen that can beconsumed by the body per unit time. A variety ofmethods are used for blood manipulation, from bloodtransfusions, administration of Erythropoietin (EPO), toliving or training at high altitudes. It is said thatBincreasing the oxygen transport capacity of the exercis-ing skeletal muscles, either by means of training ordoping, is the most powerful tool for improving athleticperformance in aerobic sports.^ ([50], p. 837) The ob-jective is to deliver more oxygen to muscles in order tosignificantly improve time to exhaustion or race times.

Several studies have documented the enhancementbenefits of blood manipulation. Three weeks of living

and training at 1380 m and simulated altitude exposureat 3000 m can improve time to exhaustion by 9% incomparison to baseline [51]. It has previously beenobserved that autologous blood transfusion improved10,000 m running performance by approximately oneminute in six highly trained male distance runners [52].Williams et al. [53] demonstrated that autologous bloodtransfusion improved 5-mile treadmill run times by 44 s,with reduced self-reported perceived exertion. A signif-icant fall has also been demonstrated in the race times ofcross-country skiers. The significantly increased perfor-mance was observed both 3 h and 14 days after reinfu-sion [54]. Following EPO administration, trained sub-jects maintained a faster pace throughout a 3000 m timetrial compared to baseline, with running performanceimproved by approximately 6% [55].

While there are other factors which can be critical toperformance, such as anaerobic thresholds, exerciseand resource efficiency, and athletes still have to trainhard, be fit and fast in order to win, the link betweenoxygen carrying capacity and endurance performanceis clearly established and intimately related. The me-chanical model of enhancement refers to such directcausal links between well determined single physiolog-ical functions and performance. Some causal links willbe more relevant to sprinting, while others to endur-ance. These mechanical links are still subject to fine-tuned interventions of dosage and range. For example,increasing muscle mass over some threshold can throwoff balance or impair agility. Our point is not primarilyabout the right balance or how many factors interact inorder to sustain performance, but rather about dissocia-ble physiological functions that are of direct relevanceto performance, which indeed do not necessarily trans-fer into performance improvements in adverse condi-tions or sub-optimal interaction. We claim that in prin-ciple one could identify a mechanical basis of perfor-mance, based on direct causal links between physio-logical functions, at least in endurance sports, all thingsbeing equal. The mechanical model does not hold forall physical sports, especially those in which neurolog-ical systems contribute in an essential manner to per-formance. For physical sports in which the practice ofa skill is central, such as balance and calculating ap-propriate body position in successful landing in gym-nastics, it would be very difficult to determine a causallink between enhancement interventions and singlephysiological mechanisms. Notwithstanding, the me-chanical model of enhancement is applicable to many

13 This suggests that motivation enhancement could potentially bemore effective than cognitive enhancement on chess performance. Asfor ethical implications, some studies document that lay people con-sider motivation enhancement as less morally wrong than cognitiveenhancement [49].

122 E. Mihailov, J. Savulescu

sports where dissociable physiological functions con-tribute to a great extent to performance.

Cognitive performance in mind sports (chess, go,bridge, draughts, xiangqi), on the other hand, cannotbe equated to relatively straightforward links betweenenhancement interventions and single neural mecha-nisms. Although it has been established that dopamine,which is modulated by methylphenidate, plays a role inreinforcement learning in response to rewards, Husainand Mehta [25] conclude that Bsimple conceptualiza-tions linking a specific neurotransmitter to a singlecognitive function are unlikely to be helpful.^ Mentalperformance, especially in cognitive sports, is morecomplex than mapping causal links between preciselocations in the brain and particular cognitive functions.Targeting single aspects of cognitive mechanisms haseven proven in some cases to have downsides. We havedescribed how Ritalin disrupts performance in a com-plex video game, Modafinil can contribute to playerslosing chess games due to running out of time andmentally calming medication can cause chess perfor-mance to collapse.

Thus, we should not think of enhancement in de-manding and specialized cognitive activities in termsof a mechanical model as illustrated by physical endur-ance performance. It might be the case that interventionswill have to work more holistically on cognitive func-tioning to prove useful, or on more complex, yet under-stood, specific domains. Future exploration could shedlight on this by comparing, for example, the effects ofpharmacological substances and brain stimulation bymagnetic fields (TMS) or electrical currents (tDCS), asthe latter could influence brain areas and theneuroplasticity of the brain as a whole.

Normative Implications: The Cassandra Problem

It is unlikely that modest improvements on basic psy-chometric tests constitute mental doping, understood asenhanced performance in demanding cognitive activi-ties. The possibility of mental doing deserves a sophis-ticated analysis which draws on interdisciplinary in-sights about how the mind works, especially in thecontext of cognitive expertise.

Another important and troubling part of the storyabout how the enhancement evidence has been misusedis its insidious moral cost. The world of sports has beenshaken by far too many examples of misconduct. There

is evidence of widespread doping across many sportsincluding athletics, tennis, football, and cycling. In con-ditions of deepening distrust, it is rational to be scepticaland devise ambitious policies to increase compliance.However, focusing only on examples of misconduct, welose sight of what has been called Bthe Cassandraproblem^ [56]. Apollo tried to seduce Cassandra, thebeautiful daughter of King Priam and of Queen Hecubaof Troy, by endowing her with the power of prophecy.When Cassandra refused him, Apollo cursed her thatnobodywould ever believe her prophecies. Even thoughshe warned the Trojans about the Greeks inside theTrojan Horse, she could not do anything because noone trusted her. For us, the Cassandra problem is thatof misplaced mistrust and unwarranted suspicion.

We are unreasonable in adopting the suspicious per-spective when we go against the evidence. While ambi-tions of having doping-free sports may seem attractive, adefault of a sceptical position is not in itself superior. AsOnora O’Neill puts it, BBlanket scepticism may soundmore sophisticated than blanket credulity, but has no moreto commend it^ (2002, p. 141). If we want to have wellplaced suspicion in mind sports, then we should not settlefor the current evidence of cognitive enhancement ingeneral. Pharmacological enhancement is characterisedby more consistent improvements in low performing in-dividuals and no significant influence on individuals withhigh performance baseline or even impairment. Further, itis characterised by improvements on some specific tasksand impairment on others, calling into question the statusof the overall enhancement. Cognitive complexity con-sists in developing and sustaining specialized cognitiveabilities, which are relatively independent from generalthought processes, usually targeted by pharmacologicalcognitive enhancement. Cognitive performance does nothave a mechanical basis of causal links between enhance-ment interventions and single neural mechanisms. Devis-ing social policy without an adequate understanding ofcognitive complexity is unwarranted and could lead tomisplaced mistrust, and consequently to an artificial cul-ture of suspicion, as appears to have happened in theIvanchuk case.

It could be objected that trust may also beundermined by not implementing a ban.14 After all,cognitive enhancers mightwork and chess players don’tneed them. However, our argument does not imply thata policy banning cognitive enhancers cannot be overall

14 We thank one anonymous referee for this objection.

Social Policy and Cognitive Enhancement: Lessons from Chess 123

justified, nor that its costs could not in principle becompensated against greater benefits. Rather, our pointis that prohibitive policies are not morally neutral withregards to undermining trustworthiness. Many peopleintuitively believe that we promote trustworthiness byprohibiting some practices or substances. While thismay be so when good evidence is available, and policiescan be effectively policed, placing many competitorsunder suspicion in spite the evidence could underminethe game. The Cassandra problem reveals a moral costwhich cannot be ignored simply by hoping that creatinga climate of suspicion erratically will lead to greatertrustworthiness in the end.

We should consider the cost of misplaced mistrustand, consequently, look for better means of reducing it.For example, a monitoring policy could be more suit-able until we see increased usage of pharmacologicalcognitive enhancement and more evidence gathered.Our analysis suggests several standards of evidencewhich are relevant for ethical analysis to decide whethercognitive enhancement is morally permissible.

Firstly, the one study involving the use of cognitiveenhancers used amateur players, it showed modest ef-fects which were reversed on subsequent games and theeffects were no greater than caffeine. What is clearlyneeded are ecological studies of elite players over mul-tiple games, comparing these effects to caffeine andother accepted enhancement practices. If Modafiniland Ritalin are no more effective at improving chessperformance than other enhancement substances likecaffeine, there are no good reasons to ban them sincecaffeine is freely permitted and these substances are nomore dangerous than caffeine. Further relevant evidencecould be gathered by comparing thresholds betweentraditional and pharmacological enhancers, not onlybetween currently permitted and prohibited substances.In one recent overview, non-pharmacological enhance-ment practices, such as naps, were found to be as effec-tive as pharmacological enhancers such as Modafinil,methylphenidate and caffeine [57].

Secondly, if we want to get more reliable indirectevidence of enhancement in chess, we should focus onaccurate models of such activities, like storage andretrieval from long-term memory of domain specificchunks of information and how this relates to patternrecognition. Talk about cognitive enhancement effectsin general should be avoided because it encourages themisperception that enhancement effects are spillovereffects. There are various and distinctive specialized

processes that make up the panoply of cognitive exper-tise, which is considered for the most part to be unrelat-ed to general cognitive abilities. We require researchinto specific enhancement effects of different cognitiveenhancers on specialized cognitive skills or holisticeffects on specific domains. While clearly it is too higha standard to require placebo controlled randomisedtrials involving elite players at tournament, it is not toohigh a standard to improve on existing research. Placebocontrolled trials involving elite players, instead of ama-teurs, in laboratory conditions against computers wouldbe a reasonable level of evidence to require. If thisshowed greater improvements than were achieved bycaffeine and napping, then a ban on grounds ofBsignificant^ performance enhancement would bewarranted.

Conclusion: Lessons from Chess for the Regulationof Cognitive Enhancement

Chess is the paradigm of a complex integrated cognitiveactivity which can be played in highly competitive envi-ronment. It is not clear whether cognitive enhancersconfer an advantage (or disadvantage) in chess, and whatthe nature and extent of that advantage might be. There isno reason based on current evidence to infer a significantprobability of providing performance enhancement ef-fects in chess. Evidence of cognitive enhancement ingeneral is too blunt and misleading. We need, instead,to conduct ecologically valid experiments on the benefitsand risks of cognitive enhancement on accurate modelsof such activities. The current evidence documents mod-est enhancement effects at basic psychometric testsbased on models of general thought processes. However,mental performance, especially in contexts of cognitiveexpertise, is more complex than mapping causal linksbetween precise neural circuits and particular cognitivefunctions. Enhancement interventions would have to bemore sophisticated than the mechanical model, whichassumes a straightforward influence between well deter-mined single mechanisms and performance. Mental per-formance is not a monolithic concept. In contexts ofcognitive expertise, performance is based on distinctivespecialized cognitive abilities that are relatively indepen-dent from general intelligence.

Cognitive enhancement is becoming more common ineducation and might also seem attractive to professionalsto improve focus, wakefulness and performance, when

124 E. Mihailov, J. Savulescu

fatigued ([58], [59]). Duke University banned the use ofcognitive enhancement on campus.15 Such a move is, inour view, premature. We should attempt to derive evi-dence involving relevant complex tasks which are basedon accurate models of such activities. Learning and pro-fessional work are complex activities similar to chess.Wecannot expect to read off from even modest improvedperformance at a simple laboratory task, a gain in acomplex activity such as problem solving. It is necessaryto identify what are the distinctive or specialized cogni-tive abilities in academic learning and particular cognitivedemanding professions, and determine how well themodus operandi of current enhancement interventionsmatches with what the sciences of the mind tell us aboutcognitive expertise in these domains.

Ethical analysis will surely be necessary to de-cide on a rational policy on cognitive enhance-ment. But in the first instance, we need ecologi-cally valid scientific research into the nature andmagnitude of effect on complex cognitive tasks.The lesson from chess is do not jump the gun.As in chess, think before you move.

Acknowledgements We are grateful to the reviewers of thisjournal for comments which helped to improve the manuscript.We also thank Cristina Voinea and Constantin Vică for helpfuldiscussions. This work was supported by the Wellcome Trust[WT203132/Z/16/Z], and by a grant of the Romanian NationalAuthority for Scientific Research and Innovation, CNCS –UEFISCDI, Project Number PN-II-RU-TE-2014-4-1846.

Open Access This article is distributed under the terms of theCreative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestrict-ed use, distribution, and reproduction in any medium, providedyou give appropriate credit to the original author(s) and the source,provide a link to the Creative Commons license, and indicate ifchanges were made.

References

1. World Anti-Doping Agency. 2014. Anti-Doping RuleViolations (ADRVs) Report.

2. Kayser, B., A.Mauron, andA.Miah. 2007. Current anti-dopingpolicy: a critical appraisal. BMC Medical Ethics 8 (1): 2.

3. Savulescu, J., R. ter Meulen, and G. Kahane, eds. 2011.Enhancing human capacities. Oxford: Wiley-Blackwell.

4. Randall, D.C., J.M. Shneerson, and S.E. File. 2005. Cognitiveeffects of modafinil in student volunteers may depend on IQ.Pharmacology Biochemistry and Behavior 82 (1): 133–139.

5. Turner, D.C., T.W. Robbins, L. Clark, A.R. Aron, J. Dowson,and B.J. Sahakian. 2003. Cognitive enhancing effects ofmodafinil in healthy volunteers. Psychopharmacology 165(3): 260–269.

6. Baranski, J.V., R. Pigeau, P. Dinich, and I. Jacobs. 2004.Effects of modafinil on cognitive and meta-cognitive perfor-mance. Human Psychopharmacology: Clinical andExperimental 19 (5): 323–332.

7. Elliott, R., B.J. Sahakian, K. Matthews, A. Bannerjea, J.Rimmer, and T.W. Robbins. 1997. Effects of methylphenidateon spatial working memory and planning in healthy youngadults. Psychopharmacology 131 (2): 196–206.

8. Linssen, A.M.W., E.F.P.M. Vuurman, A. Sambeth, and W.J.Riedel. 2012. Methylphenidate produces selective enhance-ment of declarative memory consolidation in healthy volun-teers. Psychopharmacology 221 (4): 611–619.

9. Mehta, M.A., A.M. Owen, B.J. Sahakian, N.Mavaddat, J.D.Pickard, and T.W. Robbins. 2000. Methylphenidate en-hances working memory by modulating discrete frontaland parietal lobe regions in the human brain. The Journalof Neuroscience 20 (6): 1–6.

10. Farah, M.J. 2015. The Unknowns of CognitiveEnhancement. Science 350 (6259): 379–380.

11. Franke, A.G., P. Gränsmark, A. Agricola, K. Schühle,T. Rommel, A. Sebastian, and C. Ruckes. 2017.Methylphenidate, modafinil, and caffeine for cognitiveenhancement in chess: A double-blind, randomisedcontrolled trial. European Neuropsychopharmacology27 (3): 248–260.

12. Mancuso, L.E., I.P. Ilieva, R.H. Hamilton, and M.J. Farah.2016. Does transcranial direct current stimulation improvehealthy working memory?: A meta-analytic review. Journalof Cognitive Neuroscience 28 (8): 1063–1089.

13. Racine, E., and C. Forlini. 2010. Cognitive enhancement,lifestyle choice or misuse of prescription drugs?Neuroethics3 (1): 1–4.

14. Fukuyama, F. 2002. Our posthuman future: Consequencesof the biotechnology revolution. New York: Farrar, Strausand Giroux.

15. Habermas, J. 2003. The Future of Human Nature.Cambridge: Cambridge Polity Press.

16. Sandel, M.J. 2007. The case against perfection. Cambrige:Harvard University Press.

17. Kahane, G., and J. Savulescu. 2015. Normal human varia-tion: refocussing the enhancement debate. Bioethics 29 (2):133–143.

18. Maslen, H., N. Faulmüller, and J. Savulescu. 2014.Pharmacological cognitive enhancement—how neuroscien-tific research could advance ethical debate. Frontiers inSystems Neuroscience 8: 107.

19. Repantis, D., P. Schlattmann, O. Laisney, and I. Heuser. 2010.Modafinil and methylphenidate for neuroenhancement inhealthy individuals: a systematic review. PharmacologicalResearch 62 (3): 187–206.

20. Farah, M.J., M.E. Smith, I. Ilieva, and R.H. Hamilton. 2014.Cognitive enhancement. Wiley Interdisciplinary Reviews:Cognitive Science 5 (1): 95–103.

21. Fond, G., J.A. Micoulaud-Franchi, A. Macgregor, R.Richieri, S. Miot, R. Lopez, and D. Repantis. 2015.Neuroenhancement in healthy adults, part I: pharmaceutical

15 The policy can be consulted here: https://studentaffairs.duke.edu/conduct/z-policies/academic-dishonesty

Social Policy and Cognitive Enhancement: Lessons from Chess 125

cognitive enhancement: a systematic review. Journal ofClinical Research & Bioethics 6 (2): 1.

22 . Repant i s , D. (2013) . Psychopharmacologica lneuroenhancement: evidence on safety and efficacy.Cognitive Enhancement (pp. 29–38). Dordrecht: Springer.

23. Battleday, R.M., andA.K. Brem. 2015.Modafinil for cognitiveneuroenhancement in healthy non-sleep-deprived subjects: asystematic review. European Neuropsychopharmacology 25(11): 1865–1881.

24. Finke, K., C.M. Dodds, P. Bublak, R. Regenthal, F. Baumann,T. Manly, and U. Müller. 2010. Effects of modafinil andmethylphenidate on visual attention capacity: a TVA-basedstudy. Psychopharmacology 210 (3): 317–329.

25. Husain, M., and M.A. Mehta. 2011. Cognitive enhancementby drugs in health and disease. Trends in Cognitive Sciences15 (1): 28–36.

26. Müller, U., N. Steffenhagen, R. Regenthal, and P. Bublak.2004. Effects of modafinil on working memory processes inhumans. Psychopharmacology 177 (1–2): 161–169.

27. Batejat, D.M., and D.P. Lagarde. 1999. Naps and modafinilas countermeasures for the effects of sleep deprivation oncognitive performance. Aviation, Space, and EnvironmentalMedicine 70 (5): 493–498.

28. Meeuwsen, T., I.J. Hendriksen, and M. Holewijn. 2001.Training-induced increases in sea-level performance are en-hanced by acute intermittent hypobaric hypoxia. EuropeanJournal of Applied Physiology 84 (4): 283–290.

29. De Hon, O.M., and F. Hartgens. 2000.Mind sports and doping- An investigation of pharmacological substances that mayenhance performance in mind sports. Rotterdam: NeCeDo.

30. Rogers, R.D., A.J. Blackshaw, H.C. Middleton, K. Matthews,K. Hawtin, C. Crowley, and T.W. Robbins. 1999. Tryptophandepletion impairs stimulus-reward learning while methylpheni-date disrupts attentional control in healthy young adults: impli-cations for the monoaminergic basis of impulsive behaviour.Psychopharmacology 146 (4): 482–491.

31. Schroeder, S.R., K. Mann-Koepke, C.T. Gualtieri, D.A.Eckerman, and G.R. Breese. 1987. Methylphenidate affectsstrategic choice behavior in normal adult humans.Pharmacology Biochemistry and Behavior 28 (2): 213–217.

32. Advokat, C. 2010. What are the cognitive effects of stimu-lant medications? Emphasis on adults with attention-deficit/hyperactivity disorder (ADHD). Neuroscience &Biobehavioral Reviews 34 (8): 1256–1266.

33. Gobet, F., G. Campitelli, and A.J. Waters. 2002. Rise ofhuman intelligence: Comments on Howard (1999).Intelligence 30 (4): 303–311.

34. Chase, W.G., and H.A. Simon. 1973. Perception in chess.Cognitive Psychology 4 (1): 55–81.

35. Smith, L.C., and A.A. Hartley. 1990. The game of bridge asan exercise in working memory and reasoning. Journal ofGerontology 45 (6): 233–238.

36. Frydman, M., and R. Lynn. 1992. The general intelligenceand spatial abilities of gifted young Belgian chess players.British Journal of Psychology 83 (2): 233–235.

37. Grabner, R.H., E. Stern, and A.C. Neubauer. 2007.Individual differences in chess expertise: A psychometricinvestigation. Acta Psychologica 124 (3): 398–420.

38. Chassy, P., and F. Gobet. 2011. Measuring chess experts’single-use sequence knowledge: An archival study of depar-ture from ‘theoretical’ openings. PLoS One 6 (11): e26692.

39. Hollinsky, C., G. Maresch, M. Hiller, P. Kohlberger, and C.Bieglmayer. 1997. Beeinflußt körperliche FitneßdieLeistungsfähigkeit von Ranglistenschachspielern. Ö J.Sportmed. 27: 51–59.

40. Pfleger, H., K. Stocker, H. Pabst, and G. Haralambie. 1980.Sports medical examination of top class chess players (au-thor ’s t ransl) . MMW, Münchener MedizinischeWochenschrift 122 (28): 1041–1044.

41. Fichte, J. G. 1848/1987. The Vocation of Man. trans. PeterPreuss. Indianapolis: Hackett Publishing.

42. Gazzaniga, M.S., and J.E. LeDoux. 1978. The integratedmind. Boston: Springer.

43. Fox, K.C., N.S. Fitz, and P.B. Reiner. 2017. The multiplicityof memory enhancement: Practical and ethical implicationsof the diverse neural substrates underlying human memorysystems. Neuroethics 10 (3):375–388.

44. De Groot, A.D. 1978. Thought and choice in chess. Berlin:Walter de Gruyter.

45. Ericsson, K.A., and A.C. Lehmann. 1996. Expert and excep-tional performance: Evidence of maximal adaptation to taskconstraints. Annual Review of Psychology 47 (1): 273–305.

46. Ericsson, K.A. 2005. Recent advances in expertise research:A commentary on the contributions to the special issue.Applied Cognitive Psychology 19 (2): 233–241.

47. Rikers, R.M., and F. Paas. 2005. Recent advances in expertiseresearch. Applied Cognitive Psychology 19 (2): 145–149.

48. Ramon, M., S. Miellet, A.M. Dzieciol, B.N. Konrad, M.Dresler, and R. Caldara. 2016. Super-Memorizers Are NotSuper-Recognizers. PLoS One 11 (3): e0150972.

49. Faber, N.S., T. Douglas, F. Heise, and M. Hewstone. 2015.Cognitive enhancement and motivation enhancement: Anempirical comparison of intuitive judgments. AJOBNeuroscience 6 (1): 18–20.

50. Segura, J., and C. Lundby. 2014. Blood doping: potential ofblood and urine sampling to detect autologous transfusion.British Journal of Sports Medicine 48 (10): 837–841.

51. Carr, A.J., P.U. Saunders, B.S. Vallance, L.A. Garvican-Lewis,and C.J. Gore. 2015. Increased hypoxic dose after training atlow altitude with 9h per night at 3000m normobaric hypoxia.Journal of sports science & medicine 14 (4): 776.

52. Brien, A.J., and T.L. Simon. 1987. The Effects of Red BloodCell Infusion on 10-km Race Time. JAMA 257 (20): 2761–2765.

53. Williams, M.H., S. Wesseldine, T. Somma, and R. Schuster.1981. The effect of induced erythrocythemia upon 5-miletreadmill run time. Medicine and Science in Sports andExercise 13 (3): 169-175.

54. Berglund, B., and P. Hemmingson. 1987. Effect of reinfu-sion of autologous blood on exercise performance in cross-country skiers. International Journal of Sports Medicine 8(3): 231–233.

55. Durussel, J., E. Daskalaki, M. Anderson, T. Chatterji, D.H.Wondimu, N. Padmanabhan, and Y.P. Pitsiladis. 2013.Haemoglobin mass and running time trial performance afterrecombinant human erythropoietin administration in trainedmen. PLoS One 8 (2): e56151.

56. O’Neill, O. 2002. Autonomy and trust in bioethics.Cambridge: Cambridge University Press.

57. Caviola, L., and N.S. Faber. 2015. Pills or push-ups?Effectiveness and public perception of pharmacological

126 E. Mihailov, J. Savulescu

and non-pharmacological cognitive enhancement. Frontiersin Psychology 6: 1852.

58. Sahakian, B.J., A.B. Bruhl, J. Cook, C. Killikelly, G.Savulich, T. Piercy, and P.B. Jones. 2015. The impact ofneuroscience on society: cognitive enhancement in

neuropsychiatric disorders and in healthy people.Philosophical Transactions of the Royal Society B 370(1677): 20140214.

59. Nicholson, P.J., Mayho, G., and Sharp C. 2015. Cognitiveenhancing drugs and the workplace. BMA. London.

Social Policy and Cognitive Enhancement: Lessons from Chess 127