Wesleyan University

History Teaching and Philosophy of History

Author(s): P. J. Lee
Source: History and Theory, Vol. 22, No. 4, Beiheft 22: The Philosophy of History Teaching
(Dec., 1983), pp. 19-49
Published by: Wiley for Wesleyan University
Stable URL: http://www.jstor.org/stable/2505214 .
Accessed: 22/06/2014 08:37

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Wiley and Wesleyan University are collaborating with JSTOR to digitize, preserve and extend access to History
and Theory.

http://www.jstor.org

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORYTEACHINGAND PHILOSOPHYOF HISTORY

P. J. LEE

It would probablybe fair to say that most historyteachersin Britainstill re-

gard philosophyof history(if they give it a thought)as an alien, even preten-
tious, activity,irrelevantto theirpracticalconcerns.Thisis partlyexplainedby
the attitudeshown towardphilosophyof historyby many professionalhisto-
riansin the universities,whichis characteristicallya mixtureof condescension
and suspicion;and it partlyderivesfrom the fact that teachersin the under-re-
sourced British school system have little time for reflection.In the circum-
stancesit is not surprisingthat thereis a long traditionthat teachinghistoryis
self-evidentlyhandingon to childrenin simplifiedform whateveraccount of
the past (or of certainpassagesof it) has beengivenby professionalhistorians.
The assumptionsupon which such a view rests are by no means necessarily
naiveand incoherent,but theyareoften unexaminedby thosewho makethem.
The practiceswhichthis traditionhas supportedhaveincludedsomeintelligent
and inspiredteaching,but more often have led to childrenlearninghistorical
informationratheras one might learn railwaytimetablesor telephonenum-
bers, or (marginallyless futile) as one might learn the principalproductsof
differentnations.
This traditiondoes not go unchallenged.For most of this centurytherehas
been a continuouscurrentof oppositionto the conceptionof historyteaching
as imparting"the facts," an opposition founded in the first instance on an
awarenessof the importanceof evidencein history,and on an intuitiveappeal
to somethinglike the "goodgrounds"criterionof knowledge.In recentyears
this kindof view has been strengthenedby the argumentsof Paul Hirst(which
elucidatethe notion of the developmentof mind in termsof the acquisitionof
"formsof knowledge")and by the claims of Brunerthat the "centralideas,"
"concepts,"or "fundamentals" of a subjectmay be taughtin some honestform
at school level. Togetherthesetwo strandsof thoughthaveled to increasedin-
1
terestin the natureof historyas a discipline,and argumentsas to what should
or could be taught. Otherinfluencesworkingin the samedirectionhave been a

1. P. H. Hirst, "Liberal Education and the Nature of Knowledge" in Philosophical Analysis

and Education, ed. R. D. Archambault (London, 1965); Jerome Bruner, The Process of Educa-
tion (New York, 1960).

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
20 P. J. LEE

profounddissatisfactionwith public examinationsand assessmentin history,

and an interestin taxonomiesof a Bloomian kind claimingto pick out cog-
nitiveskillsand abilities.All this led to easy dichotomiesbetween"traditional"
and "new"historyteaching, characteristically juxtaposingfacts and evidence
as mutually exclusive categories, and (often very unfairly) damning all "tradi-
tional"history teaching as meaninglessrote learning.But it also opened up
much more seriousdebate, and in particularforced those concernedwith his-
tory in educationto considermorecarefullywhatexactlyis involvedin the dis-
ciplineof history. One of the firstbooks to take analyticalphilosophyof his-
tory seriouslyas a guideto issuesunderlyingthe teachingof historywas W. H.
Burston'sPrinciplesof History Teaching,which was much influencedby W.
H. Walsh'sIntroductionto Philosophy of History.2 Burston examinedthe
concepts of fact and evidencein history, and attackedthe assumptionthat
childrencould be put into directcontact with historicalfacts or with the past
itself throughrelics, ancientbuildings,and so on. He developedan accountof
historicalexplanationbasedon Walsh'searlyelucidationof "colligation,"and
triedto deducefrom it broad consequencesfor the arrangementof eventsfor
teaching. He characterizedand examineddifferentkinds of history syllabus
(chronological,line of development,detailed "patch,""concept-based"),ex-
posingsome of the assumptionsbehindthem. Burston'sworkprecededgeneral
interestherein the ideasof Hirstand Bruner,and in some waysis still the most
rigorous(if idiosyncratic)treatmentof the problems:it defiesany attemptto
categorize it as advocating "traditional"or "new"history in schools, and
standsas a warningagainsttoo simplemindeddemarcationsand dichotomies.
It is not easy to say what role philosophyof history(whether"analytical"or
"speculative") has so far playedin the debateabouthistoryteaching.In the na-
tureof thingsmost of those engagedin the debatehavebeentrainedin history,
and even those who (like the presentauthor)believethat many of the impor-
tant questionsare philosophicalratherthan historicalhave had little formal
trainingin philosophy.At presentphilosophyof history seems to be drawn
into the argumentin two (closely related)ways: first, in suggestingwhat the
structureof the disciplinemight be; and second, in providingan analytical
basis for investigationof the "thinkingtasks"confrontingchildren.Both these
contributionshave been developedin the contextof discussionof the aims of
history teachingand the constructionof history syllabuses;in researchinto
children'sthinking;and in the design and justificationof new teachingstrat-
egies in the classroom.

2. W. H. Burston,Principlesof History Teaching(London, 1963)and (revisededition)1972.

W. H. Walsh, Introduction to Philosophy of History (London, 1951) and (revised edition) 1967.
Walshlecturedinformallyto groupsof Burston'sstudentsat the Universityof LondonInstituteof
Education;anotherimportantinfluencewas MichaelOakeshott,who remaineda friendof Bur-
ston'suntilthe latter'sdeathin 1981.Burstonby no meansacceptedall Oakeshott'sviewson his-
and
tory, but had a very high regardfor him, and found the distinctionbetweenthe "practical"
"historical"past particularlyhelpfulin connectionwith education.(See note 4.)

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 21
II

The work of both Brunerand Hirst suggestedto historyteachersthat history

mighthave its own structure,but both writersleft sufficientroom for a range
of interpretationas to where this structurewas to be located. What Bruner
counts as the centralconceptsor the fundamentalideas of a disciplineis not
entirelyclear even in the sciencesfrom which most of his examplescome: in
historythere is simplyconfusion. Concepts,key statementsof fact, heuristic
devices,or strategies:all seem to be offered.Paul Hirst'saccountof the devel-
opment of mind and the forms of knowledgeis much more precise, but al-
thoughit offersa rangeof criteriafor demarcationof the forms, it does not in-
dicate how many must be met for a "subject"to count as a genuineautono-
mous form of knowledge,or whethersome are more importantthan others.
Moreoverthe criteriathemselves-logically distinctconcepts,specifickindsof
conceptualstructuresand statements,characteristictests for truth and valid-
ity-are not absolutelyclear: can one for exampledetach conceptsfrom the
structuresin which they are embeddedand still decidewhetherin relevantre-
spects (whateverthose may be) they are logicallydistinct?It is hardlysurpris-
ing, then, that in Hirst'sdifferentexpositionshistorysometimesappearsas an
exampleof an autonomousform, and sometimesas partof the socialsciences.
(And of courseHirst is not the firstto find problemshere.) As a resultof the
opennessof these suggestionsabout the structureof a discipline,and the rela-
tive lack of interestof eitherBruneror Hirst in history, differentlines of ap-
proach have appearedamong the Britishadvocatesof new ways of teaching
history.A schematicsurveyof threeof the most popularof theseapproachesis
given in this section.3
(1) Teachinghistory is a matterof handingon substantive
historicalconcepts
This approachis a well establishedone in Britishschools. Syllabusesare con-
structedon the basis of conceptslike "revolution,""democracy,""industrial-
ization,"or "immigration." Withinmoreovertlychronologicalsyllabuses,cer-
tain conceptsare pulled out and offeredas the centralpoint of teachinghis-
tory. Thisapproachseemsto promiseescapefromthe dominationof "merein-
formation,"and from the problemsof selectingfrom the huge rangeof con-
tent available.But despitethe attractionof the strategythereare obviousdiffi-
cultiesin tryingto find the centralstructureof historyin the substantivecon-
cepts it employs. I shall mentionjust three.
First, the questionarisesas to the basis upon whichsuch conceptsare to be
chosen. Questionsof this kind push the debateback immediatelyto questions
about the selectionof content, and in the end about the objectivityof history.
(Therehas been no attemptto discusswhetherthereis any sensein whichsome
3. In what follows I deliberatelyleave open the clusterof notions at issue: structure,central
ideas,concepts,fundamentals,and others,althoughit seemsto me that at the veryleastconcepts
(substantiveor second-order),statements,laws,lessons,andproceduresoughtto be distinguished.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
22 P. J. LEE

substantivehistoricalconceptscould be especiallyclearexamplesof a genre.)

Any examinationof objectivityhas tendedto employ simplepolarizationsof
clustersof notions too readilytreatedas equivalent:"subjective,""relative,"
"probable,"and "biased"are opposed to "objective,""certain,""absolute,"
and "unbiased."It is not askedwhetherobjectivityis to be predicatedof par-
ticularhistoricalworks, historiansas individuals,or the form of knowledgeit-
self. Attemptsto avoid the issue of historicalcriteriaof selectionby invoking
educationalcriteriacan in turn lead to somethingvery close to Oakeshott's
"practicalpast,"in which historyis organizedto fit some practicalpresentin-
terest.4Of coursethe conceptof interestneed not have such narrowconnota-
tions. If it makessenseto talk of humanbeingssharinga commonform of life
at some level, thereis perhapsthe possibilityof intersubjectiveagreementas to
a wider concept of human interests,which in turn might offera (precarious)
foothold for the conceptof intrinsicimportancein history.(Connectionswith
a Wittgensteinianline of argumenthere would have to confront the ambigu-
ities of Wittgenstein'sremarkson "formsof life.") Sucha case mightbe devel-
oped to providecriteriafor selectingparticularrangesof substantiveconcepts
as a basis for historysyllabuses,althougha more interestingpossibilitywould
be to providea rationalbasisfor the selectionof particularpassagesof history.
Theremight also be an argumentfrom the natureof the disciplineitself. The
activityof historypresupposesat leasta minimalconcernfor freedomto assert
what the evidenceleadsus to believe(for ourselves,and for others);and also a
conceptionof man as rational(as opposedto arational).It presupposesequal-
ity of treatmentof, and respectfor, personsas sourcesof arguments.In these
circumstancesit seems unreasonablewithout powerful countervailingargu-
ments (and the onus is on those who would disagreeto producesuch argu-
ments)to deny that past changesin the fortunesof freedom,equality,respect
for persons, and the developmentof rationalityin humansocietiesare of in-
trinsichistoricalimportance.There is no space here to pursuethese specula-
tions; the point is that underlyingclaimsto base school history syllabuseson
substantiveconceptsare questionswhich have not been sufficientlydiscussed
in philosophyof history, let alone by those concernedwith historyin educa-
tion.5
A secondquestionfor the substantiveconceptapproachis whetherthereare
4. MichaelOakeshott,Experience and Its Modes (Cambridge,England,1933);and"OntheAc-
tivityof Beingan Historian,"in his Rationalism in Politics (London,1962).I do not meanto sub-
scribewithoutreservationto Oakeshott'saccountof the distinctionbetweenthe historicaland the
practicalpast, but would arguethat thereis a case to be made for a distinctionof some similar
kind. Recentaccountsby Britainand Argentinaof the "historicalorigins"of the Falklands/Mal-
vinasdisputeareilluminatinghere,not becauseof any simplebiastheydisplayor falsestatements
theymightmake,but becausethe pastorganizedin thiswayis so distortedas to be almostunintel-
ligible.Suchaccountsappearin schoolsas wellas newspapers.Butthe samecouldbe saidof many
"origins"accountsof (for example)the WelfareStateor the CommonMarketofferedin schools.
5. Thereare recentsignsin philosophyof historyof renewedinterestin thesequestions.R. F.
Atkinsonhas suggestedthat historiansoperatewith somenotion of whatis "humanlyimportant"
(Knowledge and Explanation in History [London,1978],85-86); see also F. A. Olafson, The Dia-
lectic of Action (Chicago,1979),Ch. 5, for some suggestiveremarkson historicalevaluation.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 23
any specialhistoricalconcepts.The Schools CouncilProject "History13-16"
was able to pick out only the conceptsof primaryand secondarysource. On
the face of it therearegood reasonsfor thinkingthat historicalconceptsare ei-
ther the practicalconcepts of everydaylife or theoreticalconceptsborrowed
from other disciplines(particularlythe social sciences).But then there is the
temporaldimension:history also employs the practicaland theoreticalcon-
cepts of those people and institutionsit is seeking to investigate.Historical
conceptsmay have a differentintensionand extensionat any momentof his-
toricaltime, so that withoutbeingentirelyfrivolousone mightsay that a fam-
ily-resemblanceaccount needs to bring in ancestorstoo. The immensecom-
plexitieslurkingbehind the appeal to historicalconcepts have scarcelybeen
touchedon in the debateso far.6Certainlythereis no clearview yet as to what
is specialabout the way in which conceptsso obviouslyfamiliarto other dis-
ciplines are employedin history, and hence no clear understandingof the
problemsfacingchildrenin learninghistory,and no real foundationin the ap-
peal to conceptsfor history'splace in school. (Theimportantfact that history
provides instances which give many of our concepts their concrete con-
tent-what is a fascist or a democraticstate?-whatever passagesof history
are studiedhas not been given the weightit deservesin debatesin Britain,de-
spite Kitson Clark'seloquent discussion.)7Would it matter if analysis sug-
gestedtherewereno specifichistoricalconcepts?The heroiccourseis to deny
that the issue is of any importance,but this is obviouslynot an easy step for
anyone claimingto move towardteachingthe centralideas of the discipline,
since it threatensto underminethe assumptionthat there is an autonomous
disciplinewhose centralconceptscan be taught. Hereworriesabout the status
of school history at last come to the surface.
Finally,is it a misunderstandingto thinkthat conceptsare centralto history
anyway?Mightnot historybe bettercharacterizedby stressingits concernwith
particulars,or historicalindividuals?This is an areain whichthereis a feeling
of vague uneaseratherthan debate. Many historyteachersare inclinedto as-
sert that historyis concernedwith the particularratherthan the general,but
the implicationsof this view have scarcelybeen examinedby the supportersof
the substantiveconceptsapproach.8In the book by Burstonreferredto earlier,
Walsh'sdiscussionof colligationreceiveda good deal of attention,but mainly
in the contextof explanationin which it was originallyset by Walsh. As a re-
sult Walsh'sbroadercolligatoryconcepts, whichhe characterizedas concrete
universals,are accordedlittle space. The Renaissance,the Reformation,the
IndustrialRevolution,all markedwith the definitearticle,are morelike names

6. An honorableexceptionis the paperby A. D. Edwards,"The'Languageof History'and the

Communicationof HistoricalKnowledge"in History Teachingand Historical Understanding,
ed. A. K. Dickinsonand P. J. Lee (London, 1978).
7. G. KitsonClark, The CriticalHistorian(London, 1967),5-7.
8. A. D. Edwardsdiscussesnamesand historicallanguage,and Burston,chapterIV, examines
claimsthat historyis concernedprincipallywith particulars(in the contextof a discussionof laws
and explanationin history).

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
24 P. J. LEE

of historicalindividualsthan concepts.Thereis at least a primafacie case for

arguingthat, fittedtogetherin a temporalframework,they providethe struc-
tureof historyat the substantivelevelin a way that conceptsdo not- at leastif
naturalscience is taken to offer the paradigmof a conceptualstructure,as
Brunerappearsto assume.
This arrayof problemsfor the first(substantiveconcepts)approachhas led
to two differentmoves. One is to regardhistoryas a storehouseof conceptsto
be plunderedfor the generaleducationalgoal of enlargingchildren'sconcep-
tual range and grasp, simplyignoringhistoricalcriteria.Such a strategycuts
conceptsloose from their interrelationsand connections,beggingmore ques-
tions than it answers.Dennis Gunning,for example,offersteachersa choice
betweenteachingthe FrenchRevolutionto promoteunderstandingof the Rev-
olution, with the "subsidiary" aim of extendingpupil'sgraspof such concepts
as "revolution,""discontent,"and "peasant";or alternativelysettingout with
the fundamentalaim of givingpupilsa graspof those sameconcepts,usingthe
FrenchRevolutionto do this simply"becauseit illustratesthe ideasin a colour-
ful, dramaticand interestingway."9He opts for the latter, and concludeshis
case with an argumentwhich is increasinglyfrequent:"If, in say five years'
time, the studentsdon't recalltoo muchabout the detailsof the FrenchRevo-
lution, I will not mind, so long as theirgraspof the conceptsis sound."10This
sort of positionis fraughtwith difficulties,but thereis roomhereonly for three
points. First, thereis no attemptto investigatewhatis involvedin understand-
ing the suggestedconcepts:historygives conceptsconcretecontent, but not as
lists of interchangeableexamples.Graspinga concept does mean graspinga
rule, but the ruleis emptywithoutinstances,and instancesdifferradicallyover
time, often stretchingthe rule to breakingpoint. This is what makeshistorical
understandingof concepts important.Revolutionin the late eighteenthcen-
tury is not the sameas revolutionin the earlytwentiethcentury.If history"il-
lustrates"concepts, different"illustrations" will radicallyalter understanding
of those concepts.The secondpoint follows directlyfrom this. A studentwho
does not "recalltoo much about the detailsof the FrenchRevolution"will not
unequivocallyhave a "sound"graspof the conceptof revolution:his concept
will be impoverishedthroughforgettinginstances-that is, historicallyimpov-
erished-even if he can give some broaddefinition.Finally,plunderinghistory
for "illustrations"is fundamentallyunhistoricalbecause it fails to recognize
that at bottom learninghistoryis learningabout the rationalinvestigationof
the past. As G. R. Elton succinctlyput it: "Historybecomesjust an instru-
ment-and so unsubtlea one that it could readilybe replacedby the Tolkien
corpus-for makingchildrenapply conceptsby means of prescribedintellec-
tual exercises.""The firstmove then (exemplifiedin Gunning'sstrategy)ends
in an atemporalitywhich rendershistory superfluous.The second move re-

9. D. Gunning, The Teaching of History (London, 1978), 16.

10. Idem.
11. G. R. Elton, "Putting the Past Before Us": review in the Times Literary Supplement, 8 Sep-
tember 1978.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 25
treatsfrom the difficultiesinvolvedin findingthe structureof historyin its sub-
stantiveconcepts, and seeks insteadto discoverit by pickingout certainsec-
ond-orderconceptsheld to be fundamentalto the discipline.Behindthis move
is a fear that therecan be no intrinsichistoricalcriteriafor the selectionof his-
toricalcontent,and a reluctanceto invokeeducationalcriteriain case they lead
back to the first move, or to some form of social studies. Hence this second
move is in realitya breakfrom the firstapproach,and requiresseparatetreat-
ment as a distinctpoint of view.
(2) Teachinghistorymust in the end come down to developingchildren's
understandingof structuralsecond-orderconcepts
The usualchoiceof candidatesto providethe structuralbasis for the discipline
turnsout to be "evidence,""cause,""empathy,""change,"and "time."It is not
alwaysappreciatedthat these conceptsare not entirelytransparent.Examina-
tion boardsin particularhavebecomeproneto offertheircustomerssets of ob-
jectivestakingthe form of lists of conceptswhich, they claim, theirexamina-
tions test. But to be told that childrenmustmasterthe conceptsof (say)change
or time withoutany indicationof what such masterywould consistin is to be
told verylittle. In any case what is importantfor notionsof historicaltime and
change is the close connectionbetweenideas of sequenceand durationand
substantivehistoricalknowledge,which means that generalstatementsto the
efffect that children should acquire"a concept of time" do not circumvent
problemsof content selection. (What such assertionsmight do is to suggest
possiblecriteriafor content selection, but to the best of my knowledgethere
has been no attemptto discussthis.) In short, the approachwhichattemptsto
locate the structureof history in a group of fundamentalsecond-ordercon-
cepts immediatelydemandsphilosophicalinquiry.Evasion of the issue here
often leads to a move towardsprocedures(see approach(3) below) which on
the face of it can be treatedsimplyas technicalmatters,to be sorted out by
readinghistorians'commentson their craft.12
Wherethe problemsare faced head-on,as in Denis Shemilt'sexcellentwork
in evaluating and directingthe Schools Council Project "History 13-16,"
teachersare often worriedthat they are beingaskedto teachphilosophyrather
than history;they fear their own weaknessesin this area, and complainthat
historyis somehowceasingto be historyat all. Such fearsin partreflecta mis-
understandingof Shemilt'swork, and in part hint at a genuinedifficulty.The
"History 13-16" course claims "to develop concepts and skills rather than
transmita body of content,"but suchconceptsand skillsmustof coursebe de-
velopedthroughhistoricalcontent.13 The point here is that the pupil who has

12. The most popular sources of wisdom seem to be E. H. Carr, What is History? (London,
1961); G. R. Elton, The Practice of History (Sydney, 1967); and, less well known, J. H. Hexter,
The History Primer (New York, 1971). Perhaps the two most interesting and insightful works tend
to receive less attention: Marc Bloch, The Historian's Craft (Manchester, 1954); and (particularly
valuable for its reference to history in education) Kitson Clark, The Critical Historian.
13. Explorations in Teaching SCP History 13-16, ed. A. Plummer, History 13-16 Project and
Southern Regional Examinations Board, 1982.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
26 P. J. LEE

completedthe "History13-16"courseought to have a more sophisticatedun-

derstandingof notions like "cause"and "change"as they operatein history
than a pupil who has studiedhistoryin a traditionalway. Shemilt'sevaluation
study proceedson the assumptionthat a child'sconceptionsof these second-
orderconceptsindexhis understandingof history.But of coursethereis a ten-
sion still betweenunderstandingwhat happenedin some passageof history,
and understandingnotionsof cause, change,evidence,and so on; in particular
thereis a differencebetween(i) giving an explanationof why somethinghap-
pened, (ii) havinga certainconceptof cause, and (iii) givingan accountof the
concept. "Understandinghistory"overtlyinvolves(i) and (ii), but not (iii).
Two points arise here. First, it is misleadingto juxtapose"developingcon-
cepts"against"transmittingcontent."Presumablythe conceptsin question,as
well as being developedthroughcontent, are developedin orderto allow pu-
pils to ask and answerhistoricalquestionsmoresuccessfully:that is, to acquire
historicalknowledgeand understandingof particularpassagesof history-in
other words, "content."This emphasizesthe artificialityof easy distinctions
betweenthe best of traditionaland the best "new"history. In both cases the
necessaryconceptsare developedthroughsubstantivecontentand the point of
the exerciseis more proficiencywith substantivecontent. ("Transmission" of
contentis not a neutralway of describingalternativesto the "development" of
concepts.)Second, any abilitychildrenacquirein performing(iii)-giving an
account of the relevantconcepts-is not an historicalachievementbut some-
thingcloserto philosophicalanalysis.How far then can we assumethat it does
serve as an index of historicalunderstanding?Shemilt'sinterviewsin "The
Devil's Locomotive"are carriedout with perceptivesophistication,but even
the briefestglanceat the excerptsset out thereshowshow difficultthis question
is. Moreoverthe performanceof some unquestionablyoutstandinghistorians
on similartasks (exhibitedin their books on the disciplineof history)must
raise questionsas to whether(iii) accuratelyindexes(ii).
Two other problemsfor attemptsto find the structureof historyin a set of
second-orderconcepts are worth brief mention. Philosophy of history has
from time to time been dividedinto analyticaland speculativebranches.On
balancethis is a sensibledivision,but some elementsof speculativephilosophy
of history- or somethingverylike it - spill over into any analyticalaccountof
explanation,cause, and other centralideas in history. If one is inclinedto be-
lieve that our knowledgeof the past shows certaincategoriesof events to be
causallymore significantthan others, one is likelyto espousean analyticalac-
count of the conceptof causein historythat reflectsthis. (Thisis not to assert
any simplisticone-to-onerelationshipbetween- for instance- a denialof any
importantcausalrole for individualsin history,and a hostilityto methodolog-
ical individualism.But oppositionto some "methodologicalindividualisms" is
not just logical.) This kind of connectionmakes it difficultto distinguishin
Shemilt'sworkbetweenthose elementsof pupils'answerswhichattempta for-
mal accountof the conceptof cause, and those which offera substantivethe-
ory of historicalchange. Consider,for example,the subjectwho declares:"If

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 27
they (a few great men) didn't want changes . . . and want wars, there'd be no
changesin historyand it'djust go on with nothinghappening."Shemilt(legiti-
mately, and with considerablesubtlety)tries to use this to get at the formal
structureof the conceptsof changeand event:"manyfifteen-year-oldsequate
'what happens'with 'what changes'-a 'change'is seen as little more than a
'new event.' Adolescents lacking concepts of change and developmenthave
difficultyin realizingthat the storypresentedto themis selective,or that selec-
tion reflectsanythingotherthanthe intrinsicimportanceof the eventsrecorded."
But what the subject seems to be offeringis a substantiveaccount of how
changeshavein fact occurredin history,togetherwith a verydefinitenotion of
historicalselectioncriteria.Whenhe says "It'djust go on with nothinghappen-
ing" he clearly does not mean that there would be no "new events"taking
place-merely none worth recording, or at least worth characterizingas
"change."And this wouldbe becausewhatmakesthingshappenin historyof a
kind significantenoughto count as "change"just are the purposes(expressed
in action)of "greatmen."(Ask a child cominghome from school "Whathap-
penedat school today?"The answermay well be "Nothing."But this does not
mean he believesnothing happened,merelythat he well understandscertain
everydayprinciplesof selection for handlingpast events.) It is preciselythis
ambiguitybetweenformal and substantiveelementsrecordedin Shemilt'sin-
terviewswhichmakethemso importantand exciting,and perhapsthe substan-
tive elementin the end does justify his use of the responsesto index children's
historicalunderstanding,albeit in a slightly differentway from the original
claim.
The other problemin this kind of approachis the tendencyfor particular
philosophicalaccountsto begin to play a normativerole in the assessmentof
children's understanding,when those accounts are still contested. Denis
Shemilt'sdiscussionof Levels2 and 3 of children'sconstructionof historical
narrativerevealchildrensubscribingto variousforms of agentcausationremi-
niscentof Taylor'sargumentsin Action and Purpose.14 Othersubjectsappear
to be workingwith a notion somewhatsimilarto Bhaskar'scausalpowersand
tendenciesof things operatingin an "openworld"without the possibilityof
laboratoryclosure.15 Thereare responsesin whichit appearsto be deniedthat
predictionand explanationare isomorphic,and otherswhich seem to be pre-
supposingsomethinglike Mackie'sidea of a "causalfield."916 Thereis not space
to discussthese issues properlyhere:the point is that the interviewsthrowup
conceptions of cause which, while undoubtedlycrude, are not necessarily

14. R. Taylor,Action and Purpose(EnglewoodCliffs,N.J., 1966).

15. R. Bhaskar,A RealistTheoryof Science(Brighton,1978)and ThePossibilityof Naturalism
(Brighton,1979).The earlieraccountin termsof "powerfulparticulars" in R. Harr6and E. H.
Madden,CausalPowers (Oxford, 1975), also providesthe possibilityof an alternativeway of
readingsubjects'answers.
16. Thereare of courseboth hermeneuticand "transcendental realist"linesof argumentwhich
denyisomorphismbetweenpredictionandexplanation.J. L. Mackie,TheCementof the Universe
(Oxford,1974),usesthe conceptof "causalfield"in a sophisticateddefenseof a basicallyHumean
account.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
28 P. J. LEE

philosophical nonsense, and which should not be measuredtoo narrowly

againsta Humeanaccountwhich is itself not unproblematicin history."7
The unresolvedtensions of the second-orderconcept approachto defining
the structureof historyturn up again in the syllabusdescriptionof "History
13-16."Thereit is assertedthat the Enquiryin Depth "helpspupilsto under-
stand people of a differenttime and place," and the Study in Development
"helpspupilsto understanddevelopmentand changethroughtime."Theprob-
lematicrelationbetweensuch understandingand the abilityto give some sort
of account of it has alreadybeen mentioned.But the tension comes out in a
differentway too: apartfrom the extremelybroadcontentselectioncriteriaim-
plicitin the syllabusstatement,the criteriaat workin choosingwhichpassages
of history should be studied are largely pedagogicaland educationalin the
widesense-based on the "needs"of adolescentsandthe practicalexigenciesof
findingmaterialfor childrento work on. This takes us back to some of the
problemssketchedout in the discussionof the first approach-for example
whetherhistoryhas any intrinsiccriteriaof importance.The second-ordercon-
cept approachto findinga structurefor historyis a markedimprovementon
the first, based as it is on an articulatedconceptionof history as a form of
knowledge,implicitly(and sometimesexplicitly)appealingto philosophyof
historyfor its justification."History13-16"exemplifiesthese strengthsin the
clearestpossibleway. But this secondapproachdoes not offera solutionto the
question of what passagesof history childrenshould encounter,nor does it
successfullysidestepthat question.
(3) Teaching history is teaching historical skills, abilities, or procedures
The strengthof this view is that, potentiallyat least, the second-orderconcepts
in (2) are explicitlyset in the activitywhichgives them point and application.
This thirdapproach,however,is in practicea broadchurch,coveringa range
of ideas and encounteringsome severeproblems.
First, as mentionedin (2) above, discussionof historicalskills and proce-
dureshas tendedto proceedas if merelytechnicalquestionsare at issue. This
leads to an eclecticplunderingof handbooksand surveysof the craft written
by historians,frequentlyaccompaniedby a concentrationon the notion of evi-
dence at the expenseof areas like explanationwhich historianshave on the
whole fought shy of, or treatedin a dogmaticway.18The "technical"approach
to evidencehas often been very simplistic.Historicalevidenceis sometimes,
for instance,treatedas a categoryof objects, examplesof which childrenare
set to learn.Behindthis lies a failureto understandthe dependenceof evidence
upon questions,a view given credenceby some historians'discussionsof the

17. See works cited in notes 14 and 15. In philosophy of history Maurice Mandelbaum, The
Anatomy of Historical Knowledge (Baltimore, 1977), Part Two, takes a non-Humean line. Oake-
shott bans "cause" from history altogether in a way which suggests Shemilt's summary of Level 2
subjects' beliefs: "everythingis thought to be connected . . . everything is relevant and there are no
loose ends." ("The Devil's Locomotive," in this Beiheft, 7.)
18. Hexter, History Primer, exemplifies the latter.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 29
subject.19The same mistakehas led to childrenbeing asked (independentof
any historicalquestion)to classifyitems like Gibbon'sDecline and Fall of the
Roman Empire as primaryor secondarysources.Anotherresultof failureto
see the complexitiesin the relationbetweenan historian'sevidenceand whathe
is preparedto assertas fact is the tendencyto discussproblemsof evidencein
termsof verylimitedconceptslike "bias."Blankettermslike this often serveto
conceal preciselythe questions and discriminationswhich are uppermostin
historicalinvestigation.Whatis this item whichwe wish to treatas evidence-
how did it arise, and what partdid it havein the societywhichproducedit? Is
it somethingupon which a conceptof bias can get a purchaseat all (a Roman
road? a piece of pottery?)except in an extendedand analogicalsense?
Second, a questionarisesabout the relationbetweenwhat historiansdo and
what pupils can do. This is not just a questionabout children'scognitiveca-
pacities, but at least equally about the characteristicsof the discipline.The
proceduresof history do not operatein individualisticisolation, but acquire
theirformand theirefficacythroughtheirplacein a publicform of knowledge.
Muchof the advocacyof childrenbeing"theirown historians,"or "doingwhat
historiansdo," fails to appreciatethis fact. Historicalinvestigationand specific
questionsgenerallyhave their originsin the currentstate of the disciplineas
perceivedby its professionalpractitioners:history is not just the common-
sense enquiryof anyone with an antiquariancuriosity.Before any work can
get underway the historianmust know as much as possibleabout the society
he is studying;this is not a mere prudentialmaxim, but indicatesthe funda-
mentalpoint that before he can put any weightupon somethingto be treated
as evidence,the historianmust know what it is. And this entailsknowledgeof
the societywhichproducedit - historicalknowledgeand the processof histori-
cal enquirycannotbe divorced.All this createsdifficultiesfor some of the easy
claimsmadeabout childrenbecominghistoriansand producingtheirown per-
sonal versionsof history,and leads W. B. Stephensjustifiablyto insist that "to
comparethe maturework of historicalresearchwith the exercisesconducted
by children from limited and preselected material . . . is the sort of clap-trap
thatbringsthe scholarshipof educationistsinto doubt. "20Nevertheless,in a re-
cent HistoricalAssociationpamphletwhich has deservedlyhad considerable
influenceon discussionof historyteachingin Britain,PeterRogershas argued
that childrenmustbeginto learnto do whathistoriansdo, and that this entails
workingwith small-scaleproblemsfor which genuinelyrepresentativeselec-
tions of evidenceare possible.21Childrenmustlearn(withhelp)to ask histori-
cal questionsof this material,assess the evidenceavailablein a criticaland
constructiveway, and drawtheirown conclusions.Rogersis well awareof the
issueslurkingbehindhis suggestions,and basesthis approachon a preliminary

19. See Elton, Practice of History, ch. II.

20. W. B. Stephens, review of archive teaching units in Journal of the Society of Archivists 4
(1970), 84.
21. P. J. Rogers, The New History: Theory into Practice (London, 1978).

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
30 P. J. LEE

generalepistemologicalaccountderivedpartlyfrom Hirst.22He arguesthat if

childrenare to be claimedto haveknowledgetheymusthavethe sort of "know
how"(knowledgeof procedures)demandedby the "rightto be sure,"and this
can only come throughlearningto operatethe proceduresspecificto the form
of knowledgeinvolved.This invocationof somethinglike the "goodgrounds"
criterion of knowledge is applied in the context of a discussion of the
historian'sprocedureswhich does not stop at a merely technicallevel, but
startsto uncoverquestionsabout the natureof evidenceand fact in history:
this alone makesthe pamphletfar superiorto any otheradvocacyof the "New
History."Rogersis rightto stressthat learninghistorymustentailsome under-
standingof the proceduresby which historiansjustify their assertions,and
right to deny that this is an all-or-nothingbusiness.Perhapsthe most impor-
tant weaknessin his accountis an implicitunderestimateof secondarysources
derived from an understandablecontempt for most school textbooks, and
fromtoo statican interpretationof the relationshipin historybetweenfact and
evidence.Onceit has beenestablishedfromprimarysources,a fact (thatHitler
invadedthe Rhinelandfor example)can itself becomeevidencefor further(pu-
tative) facts at a "higher"level (that Hitler had a clear-cutpolicy of German
aggrandizement).Not all argumentsrequirepursuitback to primarysources,
and thereseemsno primafacie reasonwhy childrenshouldnot confrontargu-
ments at the "higher"level (as they have done in the best traditionalhistory
teaching)providedthat they have at some stage acquiredsome understanding
of what lies beneathagreed(established)facts. It is hardnot to shareRogers's
conviction that this necessitateswork with primarysources:what is not so
clearis that the bulk of school historyshouldbe organizedroundthis, or that
such firm lines should be drawn on this basis between new and traditional
history.23
The most commonversionof the approachto historyteachingemphasizing
skills, abilities,and proceduresequivocatesover even the technicalissues, let
alone the epistemologicalones, by appealingto generalized"skills"such as
"comprehension," "translation,""analysis,"and "synthesis."Evenwheresome
specifichistoricalcomponentis soughtto fleshthese out, thereis little attempt
to sort out the relationbetween(say) "analysis"as practicedby the historian
and as carriedout by the pupil, becausethe "skills"in questionare conceived
as somehowunitarythinkingskills. In its most extremeform this approachin-
vokes unitaryskillswithoutany regardto differencesbetweenformsof knowl-
edge, as if analysiswere "the same"in historyand chemistry.(This is not to
deny the validityof these categoriesat a very high level-merely to arguethat
characterizingchildren'sthinkingtasksor the objectivesof teachersin this way
leads to seriousconfusion.) The obviousadvantageof this kind of positionis

22. Idem.
23. It is misleading to talk of "higher" and "lower" levels here. We often know things at a
"higher"level without being able to say anything specific at the "lower." This both poses problems
for children and offers opportunities to teachers. (See Section III.) It is not possible here to do jus-
tice to the sophistication and care with which Rogers has worked out his proposals, but I hope I
have made it clear that despite reservations I think them of great interest and importance.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 31
that it promisesan easy answerto the challengethat historyhas no value for
modernlife. If skills are unitary,questionsabout transferscarcelyarise. One
can hardlyarguethat it is not valuablefor childrento learnto analyze,synthe-
size, and so on! The morethoughtfulproponentsof the skillsapproachtendto
breakdown or add to the high-levelcategories,supplementingthem with spe-
cial historicalskillslike "empathy"or "detectingbias."But in the end any seri-
ous attemptto get clearabout what is involvedin these must drawon workin
philosophyof history.The fundamentalmistakeis to treatcategorieslike syn-
thesisand comprehensionas the namesof tightlydefinableskillson a par with
physicalskills like ridinga bicycle, whereimprovedperformanceis relatively
uncomplicatedand can in importantrespectsbe seen, and wherethe skill in
questionis developedthroughpractice.Intellectual"skills"are not like this.
What is at issue is not the broad names of styles of thinking,but the proce-
dureswhich make historywhat it is.

III

Muchof the emphasisof the "newhistory"has been on givingchildrenexperi-

ence in handlingevidence;not until relativelyrecentlyhas therebeen a corre-
spondinginterestin historicalunderstanding,explanation,and connectedno-
tions of empathyand imagination.This sectionwill be largelytakenup with a
discussionof thesetwo latterareas, in the hope that it will be possibleto show
some of the waysin whichassumptionsabouttheseaffectargumentsabouthis-
tory teaching.The consequencesof such assumptionsare to be found in deci-
sions about centralfeaturesof the disciplineof the kind sketchedin sectionII,
in the categoriesemployedin the investigationof children'sthinkingin history,
in the developmentof teachingstrategies,and in the constructionof history
syllabuses.
As we have seen, W. H. Burston's Principles of History Teaching adopted
Walsh'saccountof colligationas a theoreticalfoundationfor "topicmethod"
in history. The basis of the topics was to be policies, explicit or implicit.
Walsh'sown early discussionof colligationtendedto run togethercolligation
in termsof policiesand colligationunder"appropriateconceptions";and Bur-
ston followed this, perhapsplacing slightlymore emphasison overt policies.
But the centralweaknessof Burston'sposition was preciselythe ambiguityat
the center of his program:what sort of thing could provide the basis for a
"topic"?If the "topics"which are to constitutethe organizingprincipleof
school history may be constructedby groupingevents in terms of anything
from the explicitrealizationof a consciousand plannedpolicyto an historian's
generalcategorizationon the basis of what best makes events intelligibleto
him, then almost anythingwill qualify as a topic, and teachinghistoryby the
topic methodbecomesas hardto avoid as speakingprose.24In any case colli-

24. See W. H. Dray, "ColligationunderAppropriateConceptions" in Substanceand Formin

History,ed. L. PompaandW. H. Dray(Edinburgh,1981)for the bestdiscussionof the subjectso
far.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
32 P. J. LEE

gation in termsof the consciousintentionsand purposeswhichare integralto

policiespresupposesthat we alreadyunderstandthe actionsto be colligated:if
we are to group events which manifestthe same purposes,we must already
knowwhatthose purposesare.25The centraltasksfacingchildrenif theyareto
understand,and teachersif they are to explain,lie one step furtherback, and
somethingmoreakin to Dray'sor G. H. von Wright'saccountof historicalex-
planationmust come first. (Morewill be said about this below.) On the other
hand colligationunderappropriateconceptionsis muchmore interesting,and
its importancefor history extendsmuch widerthan the area of explanation.
Regrettablyit has receivedlittle attentionuntil recentlyfrom philosophersof
history, and almost none from writerson historyteaching.26Even basic dis-
tinctions between appropriateconceptionsavailable to contemporariesand
those whichwerenot evenin principleapplicableby them(eitherthroughigno-
ranceof outcomesor becausethe conceptswerelacking)have not been sorted
out, and discussionof criteriaof appropriateness is conspicuouslyabsentfrom
the literatureof historyteaching.If colligatoryconceptsare in some as yet un-
specifiedway centralto the substantivestructureof history,then the "newhis-
tory" must pay them much more systematicattention. In currentteaching
"new"9 or traditionalhistorycomes"pre-chunked"; childrenareofferedalready
colligatedmaterial.27Thereare obvious reasonsfor this. Childrencan hardly
be immersedin a morass of disconnectedand perhapsunrelateddetail and
simplytold to make some sense of it: at least, not as a generalrule. Neverthe-
less, a diet of exclusivelyprecolligatedmaterialdoes little to exposethe criteria
on whichcolligationsare judgedor disputed,and if openingup the possibility
of lookingat the worldin an historicalway is one of the goals of historyteach-
ing, then some graspof those criteriais essential.Even the understandingthat
colligatoryconcepts may overlap or compete would be a worthwhilebegin-
ning. It is noticeablehow accountsof the Russianinvolvementin Afghanistan
have been colligatedin Britainundera concept of Russianaggrandizement,
eitheras the workingout of an explicitprogram,or as a manifestationof the
fundamentalassumptionsof Marxism;therehas beenlittle awarenessof alter-
nativeways of colligatingthe same events, whetherin Marxistterms(the col-
lapse of a feudal regime?)or simply under some concept of popularaspira-
tions and social crisis.28Similarlythe Falklands/Malvinascrisishas been dealt
with in the Britishpopularpressunderthe organizingconceptof Argentineag-
25. This is one reason why colligation - even in terms of policies - is not satisfactory as an ac-
count of explanation in history.
26. C. B. McCullagh, "Colligation and Classification in History," History and Theory 17
(1978), offers a brief discussion. For consideration of the concept in the context of history teaching
see Burston; D. Thompson, "Colligation and History Teaching" in Studies in the Nature and
Teaching of History, ed. W. H. Burston and D. Thompson (London, 1967); and Rogers. M.
Booth, "Inductive Thinking in History: The 14-16 Age Group" in New History, Old Problems, ed.
G. Jones and L. Ward (Swansea, 1978), makes some reference to colligation. See below.
27. This is not true without qualification of Rogers's proposals for a new history based on "rep-
resentative selections" of sources.
28. It is arguable that colligations are sometimes best analyzed as the names of historical
stories.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 33
gression, and only sketchilyelsewhereunderconceptsof colonialismand de-
colonization.The point of these examplesis not to claim that what is offered
by the press is wrong, but to indicatethat childrenwho have seldom had to
consideralternativecolligatorygroupings,let alone the critieriain terms of
which their appropriatenessmay be judged, will be relativelyhelplessin the
face of what is presentedto them. Thereis a parallelhere with the discussion
of evidencein the previoussection:decisionshave to be made about what we
wantchildrento be ableto do. Shouldtheylearnto colligatefor themselves,or
is it enough that they understandsomethingof what goes on when historians
do it? Can they do the latter without some experienceof attemptingthe
former?As usual, what seemsat firstglancea straightforwardempiricalques-
tion-can childrenperforma given intellectualtask-turns out to involve a
philosophicalone: when and how far can childrenand historiansbe said to be
doing "the same thing"?(This is why Oakeshott'shighly suggestiveanalogy
with languagein specifyingthe goals of some aspectsof education-a "reading
knowledge"or a "speakingknowledge"-is in the end only suggestive.)29
The only workersin Britainwho have attemptedempiricalinvestigationof
children'sability to make colligations are Donald Thompson and Martin
Booth. Thompson'sresultsare not yet available,but Booth has made a num-
ber of claims. This is not the place to discussBooth's researchmethodology
and the validityof the results,even wereI competentto do so; whatis interest-
ing from the standpointof this paper is how conceptualconfusions make it
hardto know what could be said at the empiricallevel. Booth'sinsistencethat
"inductivethoughtis the major part of historicalthinking,"or at least "more
characteristicof historicalenquiry"than "logicaldeductivethinking"is at the
root of his interestin colligation.30Colligationis a "synthesizingactivity."31
As
suchit is "vital,absorbing,"involvesthe "creationof personalconstructs,"and
"demandsopen minded, divergentthinking."32Against induction, which is
somehowtied up with the "affectiverealmand the changeof attitudes,"stress-
ing the " 'heart'as well as the 'head'in historicalthinking,"is set deduction,a
matterof "cold, formal logic."33Indeed"deductivethinkingis destructive,in
the sensethat it analysesand breaksdown evidence."34 The task of sortingout
these suggestedpolar pairs, and evaluatingthe degreeto which they may be
treatedas equivalentin the way Booth does, is not helpedby the fact that no
accountis offeredof induction.Nor is thereany discussionof the kindsof cri-
teria againstwhich suggestedcolligationsare to be judged, beyond the asser-
tion that abstractinductiveconceptsrepresenta higherlevel of thinkingthan
concrete inductive concepts. The claim that "concreteinductiveconceptual
thinkingis less adventurous,creativeand imaginative[thanits abstractequiva-

29. Michael Oakeshott, "The Study of 'Politics' in a University" in his Rationalism in Politics.
30. M. Booth, "Inductive Thinking," 118-119.
31. Ibid., 105.
32. Ibid., 105, 120, 106.
33. Ibid., 107, 106.
34. Ibid., 120.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
34 P. J. LEE

lent] and is tied strictlyto the observableevidence"is supportedby a varietyof

evidence,but it is neverquiteclearwhatis allowablein the way of "adventure"
and what is not.35One test had subjectsputtingpicturestogetheraccordingto
their own groupings.A pupil linking picturesas showing "blackand white
peacetalks"is classifiedas thinkingat the concretelevel, even thoughthe fact
that "talks"are going on and are aboutpeace is not strictlyobservable(con-
crete)in the pictures.36 This is presumablyinsufficiently"adventurous,creative
and imaginative,"despitethe fact thatwhatis shownin one of the picturesis in
fact whitetax gatherersin nineteenth-century Africa. Are colligatoryconcepts
(leavingaside questionsabout the appropriatenessof the term here)personal
constructsin the sense of being anythingwhichcould conceivablylink the set
of test picturestogether?Or must they be justifiablein some way beyond the
fact that they do not conflictwiththe evidenceconstruedas childrenhappento
construeit? Pupil 21 linked"anAfrican king kneelingin submissionbefore a
Britishgovernor, 1896, with photographsof school childrenhaving air-raid
drill and the Nurembergyouth rally"under the concept "discrimination."37
Booth comments"Herethe evidencehas led the pupilto establishthe 'facts'of
discrimination-against the African, the Jew and the British children.The
concepthe has inducedforms an imaginativeweb aroundthe fixedpoints he
has established."38 Pupil 9 "madea set of threephotographs- air raiddrill, the
London undergroundduringthe blitz, and the Nurembergyouth rally-and
used the word 'united'to describeit."39These examplesillustratethe way in
whichBooth'sworktreatscolligationas if it wereimpossibleto judge one sug-
gestion against another, and at the same time insists that its relationto evi-
dence is important.40Colligationis a complexbusiness,balancinghistorians'
conceptionsagainstthose of contemporaries,and appealingto considerations
involvingthe currentstate of knowledgeand disagreementamong practition-
ers of the disciplinein orderto justify any particularcolligationin termsof in-
sight and fruitfulnesswithinacceptedexplanatoryideals. Thereare analogies
betweenwhat Booth'spupilsare doingand historicalcolligation,but in the ab-
senceof any clearconceptionof eithercolligationor induction,and the lack of
any examinationof the criteriafor judgingbetweencolligations,it is hard to
assess the value of the empiricalwork.
One alternativeto colligationfrequentlyofferedas a centralelementin his-
toricalexplanationand understandingis "empathy."Therecan be few notions

35. Ibid., 112.

36. Idem. See also his "Skills, Concepts, and Attitudes: The Development of Adolescent Chil-
dren's Historical Thinking" in this Beiheft, 101-117.
37. "Inductive Thinking," 111. Booth was perhaps asking leading questions at this point.
38. Idem.
39. Ibid., 112.
40. This also raises the question whether what is going on is the selection of some elements in
common to the pictures -what Werner and Kaplan call "generalization"-or whether a category is
being chosen of which all the instances are properties (what Werner and Kaplan describe as
"pluralization"and classify at a lower level). See E. A. Peel, "Some Problems in the Psychology of
History Teaching," I and II, in Burston and Thompson, Studies.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 35
so commonlyemployedin talk about what childrenneed to be able to do in
history,and so little examined.It appearsin lists of objectivesset out by exam-
inationboards(evenif it is sometimesheldto be "unexaminable"), in contribu-
tions by teachersand lecturers to TeachingHistory (the journalof the Histori-
cal Associationdealingwith schoolhistory),and in numerousmoresubstantial
discussionsof historyin education.Testsof empathyaregivenin researchinto
children'sthinking.It is an essentialpartof the SchoolsCouncilProject"His-
tory 13-16,"in whichthe Enquiryin Depthelementof the courseis specifically
designedto promoteempathy.PeterRogers'sHistoricalAssociationpamphlet
on the new historyroundlydeclaresthat empathy"is a vital part of the recon-
structiveenterprise."'41In what follows I shall not attemptto examineparticu-
lar statementsabout empathyin the relevantliterature,but insteadsketchout
somepossiblewaysof sortingout whatseemsto me the most important(recur-
rent)problemsin connectionwith the conceptas it bears on school history.
Empathyhas associationswhichare problematicfor history,and it is often
these associationswhichfiguremost prominentlyin talk of empathyin school
history: fellow-feelings,sharedemotions, and even identificationwith some
other person or group. It is frequentlytreatedas affectiveratherthan cogni-
tive, and often as a somewhatmysteriouspower, to be set againstcold logic.
Its connectionswith understandingand explanationgo largely unexplored,
and it is sometimesdistinguishedfrom and sometimesconflatedwith imagina-
tion. In the lattercase it is commonlythoughtto be somehowa suitableprov-
ince for very young children.
Perhapsthe firststep towardsortingthings out would be to distinguishbe-
tween empathyas a power, an achievement,a process, and a disposition. If
empathyis understoodas somethingwe use, a kind of specialfacultywhichis
eitherextra-evidential,or at least noninferential,then it is easy to conceiveof
it as a power. Throughempathywe know directlysomeoneelse'sthoughtsor
feelings.Thereis some sensein this conception,providedit is realizedthat the
absenceof consciousratiocinationdoes not mean that any mysteriouspowers
areat work. We sharea commonform of life at a verybasiclevelwith all other
humans,and at the culturallevel with those in our own society. Our knowl-
edgeof otherpeopledependsveryoften on our seeingcertainbehavioras signs
of somethingelse, in sucha way that no consciousinferenceis made, ratheras
a lion-trainermay know when he sees his charge'sears lie back (and without
havingto work througha deduction)that the animalis about to attack. Of
course we can make mistakes, or be fooled. There is no certaintyhere. But
wherewe are in directcontactwith people, we often get thingsright. Nothing
quitelike this can occurin history,wherewe do not perceive,or enjoy mutual
relationshipswith, the people we study. And if empathyis understoodto be a
special power by which feelings are not merelyknown, but shared, its rele-
vanceto historyis still less. Perhapsin everydaylife we may "catch"the feel-
ings of others,particularlywherethey are strongand we thinkthemappropri-

41. Rogers, The New History, 12.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
36 P. J. LEE

ate. But if this is understoodas somethingdirect, it cannot take place in his-

tory, wherethe necessaryrelationshipsare not available.Empathyas a special
power has no role in history.
What has been done when an historianhas achievedempathywith an indi-
vidualor a society, regardlessof how it has been done? Empathyin this sense
is an achievement:it is knowing what the agent believed, valued, felt, and
soughtto attain. It is beingin a positionto entertain(not necessarilyto share)
these beliefs, and being in a positionto considerthe impactof these emotions
(not necessarilyto feel them). In this way empathyis closely relatedto under-
standing.Empathy(as achievement)involvesunderstandingthe internalcon-
nectionsbetweenan agent'sbeliefsand goals, or betweenthe valuesand beliefs
of a social group. This understandingis ratherlike the sort of understanding
we need in contemporarylife if we are to work with someone or engage in
some activityin common with others. We have to be able to see how actions
can seem appropriateto a given goal, or how the goal itself can seem feasible
or desirableto the personwith whom we are dealing,whetheror not we agree
with his assessmentof means, or subscribeto his goal. Similarly,in historywe
have to be able to see (for example)how an anarchistarrangementof society
could look practicableor worthwhileto Bakunin,or nothinghe did will make
any sense to us. Empathyin this sense is doubly historical, because it fre-
quentlyrests on seeinghow, given what the agent had done and experienced,
he could havecome to see thingsand valuethingsas he did. Of courseempathy
is not all thereis to historicalunderstanding.We needalso to knowhow things
were(to the best of our judgment)irrespectiveof how anyoneat the time saw
the world, and this may entail our using concepts simply unavailableat the
time. But empathyas achievementremainsa necessaryconditionfor historical
understanding,even if it is not sufficient.
Empathymay also be viewed as a process. Empathyin historyis coming,
throughevidence,to know what an agentor socialgroupbelieved,whathis or
its values were, and so on. This conceptionis intelligibleonly as the task for
whichempathyin the previoussenseis the achievement.It is easy to drift into
thinkingthat empathizingis a specialmeans of findingout, which leads back
to a positionvery similarto the one in whichempathyis regardedas a power.
If empathizingis understoodin that way, the questions"Howcan we find out
what A believed or wanted?"and "How do we know what A believed or
wanted?"are neveransweredin historywith the reply:"Byempathizing."The
firsttask is carriedout by looking at evidence,and the second questionis an-
sweredby producingevidenceas justificationfor our assertions.Evidenceis
centralto empathizing,both as startingpoint and as touchstonefor any claims
to be made, and empathyas processdemandssound reasoning.This does not
rule out intuition altogether. Sometimes evidence from a wide range of
sources, not all of which can be recalledin detail, can lead one to think that
some agent'spoint of view was reallysuch and such withoutits being easy to
justify this "hunch"in any formal public way. Alternatively,the historian's
privateexperiencemay makehim aware,withoutany effortof consciousratio-

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 37
cination, of subtletieswhich another historianlacking that experiencemay
miss. This is whatTrevor-Roper,for example,claimsis missingin some recent
accountsof the originsof the SecondWorldWar.42In history, however,any
assertionmustin the end be justifiedby appealto publicevidence,if only in re-
sponse to specificcriticism.In any case, even these concessionsto intuition
clearlyassumea rationalrelationto evidence:suchexamplesare intuitiveonly
in the sense that the reasoningis not in the firstplace consciouslyelaborated.
Finally empathymay be treatedas a dispositionor propensity.What this
amountsto in history is the propensityto take into account other points of
view. Empathyis often characterizedas "affective": clearlyif the foregoingre-
marksare correctit makesat least as muchsenseto regardit as "cognitive"in-
sofar as thesecategoriesare applicableat all in any simpleway. But if empathy
as achievementis a necessarycondition of historicalunderstanding(whichI
have alreadyarguedis a widernotion, going beyond empathy),acquiringthe
dispositionto empathizemay be regardedas an essentialpart of learningto
thinkhistorically.Here thereis a link with the "affectivedomain."Such a dis-
positionis not unlikethe "rationalpassions"whichmust be acquiredin learn-
ing any discipline.We would not acceptthat anyoneproperlygraspedmathe-
maticswho had no concernfor accuracyor cogency,or that a studentwho had
no interestin truthreallyunderstoodwhathistoryis about. Perhapsit is in this
sensethat someonewith no dispositionto empathizecannot be said genuinely
to have learnedhistory. Historiansmust want to achieveempathywith their
subjects.But none of this marks off empathyfrom other aspects of history:
historiansmustwant to maketruestatements,respectevidenceand so on too.
Empathyas achievementmeans entertainingbeliefs and goals which one
neednot share,and it is the idea of entertainingpoints of view otherthan one's
own which providesthe link with historicalimagination.Wherethereis con-
clusiveevidencefor the viewsof some agent, andthe connectionbetweenthose
viewsand the agent'sactionsis clear,imaginationmay barelyenterinto things.
True,knowledgeof the agent'sbeliefsand goals mustnot remaininert;the his-
torianmust see themas bearingon the relevantaction. Wherehe cannotshare
the views of whomeverhe is studying, he must entertainthem consistently
enough to follow the reasons which give the action its intelligibility,setting
asidehis own beliefs. But whilethe schemaheremay be "Supposeone believed
this and wantedthat, what follows?"the role of imaginationis minimal,be-
cause what the agent believedis alreadyknown undera descriptionwhichties
it to the actionin question.In the light of the evidenceas to the beliefs of trade
unionleadersabout the Taff Vale decision,even an historianwhoseown views
about the role and value of unions differradicallyfrom those of his subjects
will needlittleimaginationto follow theiractionsin tryingto overturnthe con-
sequencesof the decision.But wherethe samehistorianis attemptingto under-
stand how people could see trade unions as desirableat all, it is a different
story.Even with particulareventsthereis often no conclusiveevidence,for al-

42. H. R. Trevor-Roper, "History and Imagination," Times Literary Supplement, 25 July 1980.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
38 P. J. LEE

though it may be clear how someone (or some group) saw things in general
terms, there may be no specific evidence about his (their) reaction to the partic-
ular event or situation with which the historian is concerned. Supposal, and
hence imagination, comes in here in a much more substantial way. Suppose
one had lived through such and such events, believed this, and wanted that,
how would one see the event in question? Imagination carries some implica-
tion of going beyond the evidence, at least insofar as what is conclusively given
can hardly require imagination. In history imagination (as supposal) operates
where a point of view for which there is evidence is "cashed"in terms of its im-
plications in a context for which direct confirmation may be lacking. Hence
Christopher Hill takes to task historians who "argue that the miseries of the
workers during the Industrial Revolution are the ex post facto invention of
sentimental historians."43The Hammonds, he insists, "were more gifted with
historical imagination" than their critics." Why? Because the mere description
of pay or factory working conditions is not enough. The historian must enter-
tain sets of beliefs and goals not necessarily his own, and with them a concep-
tion of what is normal or reasonable (perhaps enshrined in habits), and bring
all this to bear on his description. And in so doing he re-describes it. This is
imagination as supposal: it involves seeing what (in a sense) we already
know-that people earned more money for the same hours-in a different
light. Hill says at one point "We may suppose, looking back after two cen-
turies, that the imposition of labour discipline was easily accepted because its
necessity is obvious to us."45But in so supposing, Hill argues, we would be
wrong. The mistake would have been to make twentieth-century suppositions,
which is in reality not to make any suppositions (engage our historical imagi-
nation) at all, but to go on thinking in the ways to which we have become ac-
customed. This is a mistake because the evidence does not allow us to say that
such ways of thinking were normal in the eighteenth century. It amounts to a
failure of historical imagination. But it also shows that imagination is tied in
history to evidence, and is not free floating and viciously subjective.
The task of imagination as supposal is to change our point of view: then
everything looks different, and we can begin to understand.
We look backwithtwentieth-century preconceptions.After two hundredyearsof trade
unionstruggle,wagelabourhas won a respectedandself-respectingpositionin the com-
munity.But if we approachwage labourfrom the seventeenthcentury,as men in fact
did, we recallthat the levellersthoughtwage-labourershad forfeitedtheirbirthrightas
freebornEnglishmen,and should not be allowedto vote. This traditionalattitude,to-
getherwith the fact that many factorieslooked like workhouses,and wereoften con-
sciouslymodelledon them . .. may help us to understandwhy independentcraftsmen
clung so hard to economicallyuntenablepositions.46
Of course there is much more to imagination in history than this, but perhaps

43. Christopher Hill, Reformation to Industrial Revolution (London, 1967), 260.

44. Idem.
45. Ibid., 261.
46. Idem.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 39
enough has been said to indicateits close connectionwith empathy.47Where
genuinehistoricalempathyhas been achieved,it cannotremaininert. In imag-
ination as supposalhistorianscash points of view entertainedon the basis of
evidence,followingup the consequencesof those points of view, wherethere
may be no directevidenceas to what the consequenceswere. Imaginationis
thus a criterionof empathy,and so of understanding.48
In discussingempathyand imaginationI have repeatedlytalked about the
historianentertainingbeliefs and goals which he need not share. The point
hereis that empathydoes not requireagreementwith or acceptanceof beliefs,
values, purposes,goals, or policies. What then is sympathy?Surelyit is usual
to praisean historianwho displayssympathyfor his subjects,or perhapsfor
his period?In everydaylife havingsympathyfor someoneis sometimesa mat-
ter of sharingthat person'sfeelingsand emotions,but it can also be the recog-
nition of those feelingsas appropriatein some way. Often the appropriateness
is recognizedonly in the limitedsensethat the emotionsfit the circumstancesas
seen by the agent: thus althoughwe might recognizethat given his view, his
emotions are appropriate,we may want to raise questionsas to whetheror
how far his accountof the circumstancesis correct.Hencewe say "I can sym-
pathize with your feelings, but . . ." or even "I sympathize with your position,
but...." In historysympathyis morea matterof recognizingthe appropriate-
ness of feelingsthan sharingthem. Historicalsympathymay implyagreement
at a very basic level with the assumptionsand presuppositionsof some period
(insofaras it makessenseto talk like this). An historianmay findthat he tends
to agreewith the moraltenets, style of life, and idealsof some social groupat
some particulartime, perhapssharingwith them a conceptionof what man is
or mightbecome. But suchagreementis not necessaryfor good history,and if
it is more than loose and general,it may turn into merepartiality:agreement
with one set of contemporaryassumptionsimpliesdisagreementwith others.
Talk of historianshavingsympathyfor whatthey are studying,whenintended
as praise,is in partan indicationof whatthey do not do - namelyjump out of
theiraccountsto show their audiencehow misguided,stupid, or wickedsome
way of doing thingswas. Morepositively,it impliesthat becausethe historian
recognizesthe appropriatenessof the subjects'emotions, given their assump-
tions, he or she is morelikelyto be able to entertaintheirviews (includingany
emotional charge) in a consistentmanner, to have insight into their conse-
quences,and to pass all this on to the audience.
If it is misleadingto take sharedemotionsor beliefsas centralto the analysis
of historicalsympathy,then that kind of analysiswould be doublymisleading
for empathy.Sharingthe beliefs and feelingsof one's subjectscannot be cen-
tral to empathyif empathyis to be centralto history.This is not just because
of the possibleundesirabilityof such a concurrencefor impartialityand move-
47. See E. J. Furlong, Imagination (London, 1961), for what is in many ways still the most
thought-provoking discussion for those concerned with history.
48. This case is argued further in P. J. Lee, "Historical Imagination" in Learning History, ed.
A. K. Dickinson, P. J. Lee, and P. J. Rogers. Heinemann Education Books, forthcoming in 1984.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
40 P. J. LEE

ment betweenpoints of view, but becauseit is logicallyimpossiblein a great

many cases. In the first place it must be obvious that the historianscannot
sharea belief whichthey know to be false, and historicalagentsare frequently
mistakenin whatthey believe.In the secondplace,this logicalimpossibilityex-
tends into the area of emotions. Emotionshave a cognitivebasis-it is beliefs
about things which distinguishfear from hope, jealousy from anger, and so
on. Whatthis meansin historyis that I cannot(for instance)sharethe triumph
and hopes of a politicianimmediatelyafter an electionif I know that her as-
sessmentof the significanceof her victory for the future was wildly wrong.
And an historian unable to share Cromwell'sbeliefs in providencecannot
shareCromwell'semotionsduringthe 1649campaignin Ireland.The priceof
tying empathy(and even sympathy)too closely to feelingsand emotionsis to
make it a mereepiphenomenonof history(and a relatedpoint could be made
about imagination).Empathyrequiresthat the historianknows that the sub-
jects believedwhat they did with regardto facts and values, and that they felt
as they did. It does not requirethat historiansshare eitherthe beliefs or the
feelings.Whatit does demandis that they can recognizeat some level theirap-
propriatenessin their context.
The sketch given here of empathyin history is too superficial,crude, and
dogmaticto be dignifiedwith the title philosophyof history,but it will serveto
make some points illustrativeof the relationshipbetweensecond-orderques-
tions about history and issues in historyteaching.Empathyis centralto his-
tory-one might say structural,in that withoutit (as somethingachieved)his-
tory cannotbegin. In the firstplace, unlessthe point of view of its creatorscan
be discoveredand entertainedby the historian,there is no prospectof using
evidencein the way historiansmust do if historyis to be possible.At the very
least we mustknow what somethingis, whatit was meantas. The circularityin
this (empathymust itself be based on evidence)is only apparent,becausein
historyas in sciencewe cannotquestioneverythingat once- some thingsmust
be provisionallyaccepted.In the secondplace, thereoperatesin historya pre-
suppositionof rationality:that thereare internalconnectionsbetweenbeliefs,
goals, actions, and so forth whichmay be graspedby anyonepreparedto en-
tertainthem.The presuppositionis of coursedefensiblein particularcases, but
acquiringthe dispositionto achieve empathyis part of learningthis central
presupposition.49 What is requiredis not that childrencan give an accountof
empathy, but that they do not writeoff people, institutions,and waysof look-
ing at things as absurdor somehowmentallydefective.50Historyteachingis
concernedhere as elsewhereto foster the relevantrationalpassions.5'
49. Without some presupposition of this kind history as narrative is only possible in a truncated
sense, because there is no possibility of tracing intelligible connections between events as opportu-
nities, threats, or constraints, and so the backward referencing of action and of other aspects of in-
dividual and social life disappears. But it is this backward referencing which provides a basis for
historical stories. See Olafson, Dialectic ofAction and also the discussion of quasi-causal chains in
G. H. von Wright, Explanation and Understanding (London, 1971).
50. Research, experience in training teachers, and discussion with practicing teachers all suggest
that there is a strong tendency to treat people in the past as defective in this way. See below.
51. What is at issue here is not a generalized moral tolerance, or indeed tolerance in any normal

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 41
Thereare consequencesfor children'sthinkingtoo, and for our assessment
of empiricalresearch.Empathyin historyis not just a simple humanresponse
particularlysuitablefor youngerpupils.If thereis any truthin the sketchgiven
above, it is likely to prove extremelydifficultfor children.Of course the ut-
most caution is requiredin transposingsuch a sketch from maturehistoryto
historyin school. To takejust one example,it may be necessaryfor childrenat
certainages to sharethe point of view of the agent they are studying,if they
are to see its relationto action, let alone to conceiveit in its relationshipswith
other events. The developmentof empathymay even demand identification
withthe agent, somethingwhichin professionalhistorywouldbe a signof fail-
ure of imagination,and would spell disasterfor attemptsat historicalunder-
standing. Only empiricalwork can discoverwhetherthis is a step through
which pupils must pass before acquiringthe ability to move freely between
differentpointsof view. By the sametoken it is not clearhow far childrenmust
have had similarexperiencesto those past people they are studying-and in
this case, eventhe priorquestion,whichis not empirical,of whatis to count as
a "similar"experience,is virtuallyunexplored.
Whatis availableis generalpsychologicalevidenceon the widerissue of un-
derstandingotherpersons,usuallycharacterized as "empathy" by researchers.52
Muchof the work has involveda strongperceptualelement,in whichchildren
are askedto say what can be seen (of some arrayof modelsor objects)from a
differentpoint of view. Often the tests are inspiredby dissatisfactionwith Pia-
get'sgeneralaccountof egocentricityin young children,and derivedin partic-
ular from his work with the "threemountains"experiment.At the same time
there has been another strand which has concentratedon the ascriptionof
thoughts or feelings to other people, often charactersin specially devised
stories.Therehave also been experimentsin whichchildrenhave had to com-
municatecertainfeaturesof a visualarrayto a listenerunableto see what the
subjecthad in front of him.
Thereis no doubt that much of this recentresearchis highlysuggestivefor
work on children'sthinkingin history, and that the long-runimplicationsfor
historyteachingare likely to be far-reaching.But at the momentthe research
raisesmore questionsthan it answers,and a greatdeal of theoreticaland em-
piricalworkis requiredeven beforethe rightquestionscan be asked.Thereare
obviousgapsbetweenthe generalresearchand whatwouldbe neededin histor-
ically-orientedwork, and these gaps are conceptualrather than empirical.
Even in the largelyperceptualexperimentsthe distinctionmust be made be-
tween what the other personcan see, and how it is seen: Flavellsuggeststhat

sense at all. Criticism is still in order, and often necessary for understanding, as Collingwood fre-
quently insisted. The requirement is for a special, formal, and detached tolerance of the kind ap-
propriate to entertaining, not subscribing to, other points of view, and grasping the appropriate-
ness of beliefs and actions in relation to circumstances, values, and other beliefs.
52. See J. H. Flavell, "The Development of Inferences about Others," Understanding Other
Persons, ed. T. Mischel (Oxford, 1974); H. Borke, "Piaget's View of Social Interaction and the
Theoretical Construct of Empathy" in Alternatives to Piaget, ed. L. S. Siegel and C. J. Brainerd
(New York, 1978); and M. Donaldson, Children's Minds (London, 1978).

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
42 P. J. LEE

the second step is harderand comes later in a child'sdevelopment.53 This is

only to be expected,but if it is harderwith a perceptualtask, how muchharder
still will it be with an historicalone? In historyit is seldomsimplya matterof
what physicalobjects an agent (or group)could see, or how the objects were
seen, but rathera questionof what beliefs, purposes,and valueswereheld, or
what considerationswere borne in mind and what weight attachedto them.
Anotherillustrationof the way in whichresearchinto children'sthinkingin
historymust both take accountof and go beyondthe generalwork so far un-
dertaken is in connection with MargaretDonaldson's notion of "human
sense."54Her suggestionsare perceptiveand important,but she does not al-
ways clearlydistinguishbetweenthe sensea task mightbe givenby childrenas
opposed to experimenters,and the sense implicitinternallyin the task. She
rightlypoints out that what experimentersmake of a conservationtest, and
what childrenassumeis going on, may be quite different.But in some of the
tests she describes- for exampleone in whichboy dolls had to be hiddenfrom
model policemenbehind an arrayof model walls-the point about the task
makinghumansensehas a doubleimportance,becausein a test designedto get
a children'sgraspof other people'spoints of view, what the "people"actually
built into the test (the boy and policemandolls) are doing must also make
sense. Displacedvolumesor rowsof objectsdo not themselvesdo thingswhich
make sense, althoughthey figurein activitieswhich do. Boys and policemen
may figurein experimentallyimposedtasks which make sense, but may also
themselvesdo thingswhichmake (or fail to make)sense. In tests of children's
thinkingin historythis is of paramountimportance:being askedto see things
from another'spoint of view must be embodiedin a meaningfultask, but it
also involvesunderstandingwhat sense can be made of what others do, and
their purposes,intentions,and perspectiveswithin the task.
Of course there are additionalreasons why seeing things from another's
point of view is likelyto be difficultin history:the opennessand complexityof
the situationsencountered,the difficultyof specifyingin a clear-cutway what
intentionsand purposesare involvedin a particularaction, and the lack of any
mechanicalprocedurefor decidingwhat is to count as the successfulachieve-
ment of a goal are obviousexamples.Moreover,if it is understandingwhichis
at issue, it is not just the agent'spointof viewwhichchildrenmusttakeinto ac-
count, but what evidence suggests is the case regardlessof what the agent
thought. This demandsoscillationbetween(at least) two points of view.55All

53. Flavell, 95-97.

54. Donaldson, 23-25.
55. I have discussed some of these issues elsewhere: see "Explanation and Understanding in His-
tory," and A. K. Dickinson and P. J. Lee, "Understanding and Research," both in History Teach-
ing and Historical Understanding. The latter paper attempts a critique of the assumptions which
must be made if Piagetian categories for the analysis of children's thinking are to be applicable to
history. Briefly, it argues that the most important problems arise because (i) history does not have
for its subject matter experimentally manipulable objects, accessible to direct inspection, so chil-
dren cannot observe the consequences of their actions on the test material; (ii) historical questions

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 43
in all it is evidentthat any seriousinvestigationof children'sabilitiesin these
areas requiresmuch more attentionto exactly what it is that they are being
asked to do.
There are other, more immediatereasons for paying more attention to
philosophicalanalysis.Examinerswho assertthatempathyis unexaminableon
the groundsthat it is a mysterious,almostineffable,processhave simplyfailed
to sort out conceptualmisunderstandings; sucha judgmentis not an empirical
assertionabout examinationtechniques.The same is true of similarbans on
the assessmentof historicalimagination,at least in some of its forms. There
may well be technicalproblemsin producingan intervalscale for evaluating
how far pupils have performedhistoricaltasks with imagination.Even so,
when we say historianshave displayedhistoricalimaginationwe are talking
about a publicachievement,and it is not necessaryto know about their feel-
ings, privateexperiences,or images-what has gone on in their imagination.
This is trueof childrentoo. In any case, imaginationas supposalis in principle
assessable,and indeed,if it is a criterionof understanding,thereis a strongar-
gumentfor insistingthat one test of examinationand assessmentproceduresis
that they make provisionfor assessingit.
Finally,any accountof (or any set of implicitassumptionsabout) empathy
has consequencesfor teaching. Many of the problemsin historyteachingdo
not stem from a concernto pass on historicalfacts, but from a misguidedde-
terminationto do too much, and this in turncomes from a failureto stop and
sort out whatthe problemsfacingchildrenactuallyare. If vast periodsof time
are dealt with each year, the events and actions of which childrenhave to
makesenseare simplyunintelligible,and historybecomesthe absurdbehavior
of mentaldefectives.This is becausethereis no time to buildup a grasp,how-
evertentativeand simplified,of the beliefs, values,and goals of the peoplebe-
ing studied,let alone to fill in the constraintson action and behaviorimposed
by the circumstanceswithin which at any particularmoment in history they
had to work. And wherethe constraintsare not understood,any actionis pos-
sible, includingmany which on the face of it were much more likely to have
been successful(in termsof the agent'sgoals). So NevilleChamberlainshould
have met Hitler'stemperat Berchtesgadenwith a similardisplayof his own,
and laid out the Fuhreron the spot. LouisXVI shouldhaveexecutedall his op-
ponents,"beenstrongerand more forceful,"or alternativelywon overhis peo-
ple with sweet reason and the successfulprosecutionof "revolutionary" war.
The Anglo-Saxonsshould have abandonedoath-helpingand the ordeal, and
adopteda professionalpolice force and forensicscience,or at least a modern

and explanations cannot be assumed to be logically the same as those in natural science: giving rea-
sons for action is not the same as discovering regularities in natural phenomena; (iii) there is no
possibility of "all"the evidence required in a piece of history being provided for children in a test;
and (iv) the basis of Piagetian categories in propositional and class logic conceals important differ-
ences between history and natural science. Empirical work which attempts to apply Piagetian the-
ories and test procedures directly to history is hence in important ways conceptually misbegotten.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
44 P. J. LEE

conceptof evidence.56This failureof historicalagentsto see the obvioussolu-

tions to theirproblemsconfirmspupils'suspicionsthat the past is peopledby
primitives,and fits well into an idea of progressstill prevalentin an unhistori-
cal world. It renderswhole sections of the past arbitraryand unintelligible.
The tasks involvedin achievingempathyin historyare difficultand demand-
ing, and above all they need time. The syllabusimplicationsare clearenough:
historycoursesfor childrenmustbe less ambitiousin chronologicalscope, and
for some of their work at least childrenmust be given the opportunityto ex-
plore short passagesof historyin depth.
Similarconsiderationsapplyto teachingmethods.Usuallyit will be impossi-
ble for a teacherto give a sufficientlytight specificationof the circumstances
underwhichan agentactedto ruleout all alternativecoursesof action. This is
becausea graspof the acting-situationpresupposesa wide rangeof knowledge
of a past societyand how those withinit saw themselvesandthe world, andbe-
causethe assumptionschildrenmust bringto standin for theirlack of knowl-
edge will depend upon their own experience.Such assumptionsmay lead to
misapprehensions,and those misapprehensions cannotnecessarilybe defeated
in advance:a situation-descriptionwhich ruled out all possible misconcep-
tions, wrong weightingsof considerations,and so on would be indefinitely
long and detailed.Theproblemsarecompoundedwhenchildrenfail to makea
distinctionbetweenwhat the agent believedand wantedand their own views
(includinghistoricallyjustifiedbeliefs about what mightbe calledthe realities
of the agent'ssituation).57Hencethe creationof classroomopportunitiesto ex-
pose misconceptionsare of criticalimportance:what is neededis the chance
for childrento talk in ways whichbringtheirassumptionsinto the open. His-
tory gamesand simulations,drama,and role-playingmay be importanthere.
In general any task in which pupils have to predict the responseshistorical
agents or groups might be expectedto make to events, given certainbeliefs,
goals, values, and a particularpast of theirown, will be likelyto demandgen-
uinelyhistoricalimaginationand henceto test empathy(as achievement).Such
recommendationsraisethe hacklesof manyteachers:historyis concernedwith
whatdid happen,not with whatmighthavebeen. Thereis a good dealof truth
in suchstatements,and whenthey aremadein answerto certainkindsof skep-
ticismor subjectiverelativism,thereis no harmin them;but they are too sim-
ple for the matterat hand. In the firstplace, if historianseverassertnecessary
conditions for events, there is some assumptionas to how the world would
have "runon"if the necessaryconditionhad not materialized.58 This is a direct
appeal to "what would have happened if. . . ." In the second place, despite
their denial of interest in "mighthave beens," historiansare committedby
morethanimplicitlogic to suchspeculation;they arenot aboveexplicitappeal

56. Transcriptsfromcurrentresearchby the authorandAlaricDickinsonin whichsmallgroups

of childrendiscusstheseAnglo-Saxoninstitutionsindicatesthat this is almostinvariablythe first
reactionof the children(aged from eight to eighteen)undertest.
57. See Dickinsonand Lee, "Understanding and Research."
58. See Mackie,33-34.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 45
to alternativecourses of events. Considerthis passage from Elton (hardlya
frivoloushistorian):"Thequestionwhichmattersis whetherthe LutheranRe-
formationwas really necessary,or whetherthe Churchwould have been re-
newed and restored by the widespreadmovement of Christianhumanism
whichlooked to Erasmus."59 The point of this referenceto a possibleworldin
which Lutherand his "morerevolutionaryposition"did not exist is to show
that without the popular support won by Luther'sradicalstance, Christian
humanismwould have had a purelymarginaleffect.This is "whatwould have
happened if . . ." with a vengeance, but there is nothing illegitimate about it.
A similardisquietis to be found among historyteachersat the suggestion
that historicalfictionhas an importantrole in historyteaching-fiction is after
all merelyimaginary.The force of theseobjectionsis diminishedonce it is un-
derstoodthat historiansfrequentlymakestatementson the basisof evidenceat
one level, without being able to say anythingspecificabout what must have
happenedfor the generalstatementto be true. Thus we know that Maiden
Castlewas taken by the invadingRomanarmy, and we can say a little about
the weaponsused, the kinds of casualtiessuffered,and the tacticsemployed.
What we cannot say is how the many individuals involved spent their
day-who was hit or missedby slingshot,javelins, or ballistabolts, who hid,
who haranguedhis compatriots,and so on. Still less can we fit the day'sdoings
into the story of a particularlife. In consequencechildrenoften have to deal
with eventsat a level which reduceshistoryto somethinglike a formalcalcu-
lus. Writersof historicalfictionwork down from this abstractlevel, inventing
what the historian is unable to reconstruct, but this invention of the
(merely)imaginaryis basedupon evidence.Supposetherehad beenan individ-
ual with this particularset of experiences,who believedso and so, livedin such
and suchcircumstances,and took partin this (historical)seriesof events:what
might have happenedto him? The people in the story resembleideal types:
theyembodya particularconfigurationof beliefs,goals, and economicand so-
cial roles whichare at least historicallypossible,and which, given what histo-
rians have establishedon the basis of evidence,amount to a highlyprobable
combination.Idealtypesareemployedin historyanyway-Maynard's Russian
Peasant,Brenan'sAndalusianAnarchists,Soboul'sSans-Culottes-but in his-
torical fiction they performdeeds, have conversationsand personalrelation-
ships, and sufferday-to-dayevents for which therecould be no evidence,be-
causethesethingsneverhappened.Idealtypesin historyilluminateand exem-
plify widermatters,but in historicalfictionthey are given all the trappingsof
(imaginary)life.
Imaginationin historicalfiction must recognizethe constraintsimposedby
evidenceand establishedhistoricalfact, but at the same time it must meet the
ordinaryliterarycriteriaof coherence,humanpossibility,and so on. Fromthis
standpointit is also normallypart of the point of historicalfictionto evoke a
response,and the authormay activelyseek to produceemotionalreactionsin

59. G. R. Elton, Reformation Europe 1517-1559 (London, 1963), 282.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
46 P. J. LEE

his reader.The storyis usuallytold verymuch from one point of view, and in
this respectthereis rarelyany attemptto maintainan historian'sdetachment.
In children'sfiction (and perhaps adults' too?) the readeris encouragedto
identifywith the hero, whichdemandslivingin imagination.And to the extent
that the work forces the readerto accommodateto the hero'sworld, and not
merelyassimilateit to his own, it demandsimaginationas supposaltoo. Writ-
ing historicalfiction is not doing historicalresearch,and readingit is not (in
some respects)like readinghistory.But it can leadto empathyas achievement,
and an appreciationof a period, if from a narrowpoint of view, by makingit
easierto "cash"historicalknowledge.Too easy, it mightbe said:historicalun-
derstandingdemandsmovementbetweena 'rangeof points of view, some of
themneverheldby (or evencomprehensibleto) the peopleunderinvestigation.
This, of course, is one of the things that makes historydifficultfor children,
and historicalfiction, in narrowingthe rangeof viewpointsbut exploringone
or two in concretedetail, arguablyhas a muchmore importantplace in learn-
ing historythan it has generallybeen given.

IV

Assertionsabout the nature of history as a disciplineraise questionswhich,

pursuedany distance,are questionsin philosophyof history.Theyneednot in
the firstinstancebe very esotericor subtle, and may uncovera rangeof philo-
sophicalissues in one practically-orientedtask-the attempt, for example,to
sort out what is meant by evidencein history. But of course any seriousdis-
agreementhere necessitatesprogressiveclarificationand differentiationof the
questionsat issue. What is the relationbetweenevidenceand fact in history?
How far can one legitimatelytalk of historiesor even historicalstatementsbe-
ing true? This path can lead back to the fundamentalsof epistemology,but
practicalmattersare at stake all the time. The Schools CouncilProject"His-
tory 13-16" sets an evidencepaper in which examinationcandidateshave to
use unseen evidencefrom a periodthey have not studied, in the almost total
absence of context. The Project defends this on the pragmaticground that
thereis no otherway to assesspupils'skillsas opposedto theirknowledge.The
questionthen ariseswhetherit even makessenseto speakof using evidenceif
no context exists which could allow the evidenceto be interpreted.How little
context can there be before historicaljudgmenton evidencecan get off the
ground?Sometimeshistoriansmay have to makeextremelytentativeassump-
tions before they can begin to use evidence,but generallyeven wherethe con-
text of specifichistoricalknowledgeof the relevantsociety is almost nonexis-
tent, thereis a widercontext-archaeological, geographical,or anthropologi-
cal -which allowsthe work to get underway. It mightbe arguedthat the task
facing childrenbeing askedabout writtenevidenceon CardinalWolsey, when
they know nothingof Englishhistorybeforethe nineteenthcentury,is isomor-
phic with that of an historianbeginninginvestigationsinto (say) a hitherto
unknownAmericancivilization. Childrenarc forced to postulatea context,

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 47

just as the historianis. But of coursethe historian'spostulatedcontextis itself

built on evidencederivedfrom his dig, or from the way his documentcameto
light. He may not know how it fitted into a society, but he knows how it fits
into an investigation.Childrenfaced with an unseen documentfrom an un-
knownperiodare compellednot so muchto postulatea contexton the basisof
whatmightbe calledproto-historicalor investigativeevidence,but to makeas-
sumptionsabout the putative evidenceitself, and hence indirectlyabout its
context. If thereis any truthin the assertionthat childrenareonly too readyto
assumethat people in the past had the same beliefs and goals as us, but were
defectivein thinkingand judgment,thereis a dangerthat this kind of exami-
nationpaperwill activelyencouragean inadmissibleset of methodologicalpre-
suppositions.Whatchildrenareinclinedto believeis a matterfor empiricalin-
vestigation.Thereis also anotherempiricalquestionas to how much contex-
tual informationis necessaryto defeat such unwarrantedassumptions,and at
what point extrainformationbeginsto defeat the object of a "skills"test. But
in the end the questionhow far it is possibleto make sense of historicalevi-
dencewithoutcontextis an epistemological,not an empiricalone. This exam-
ple shows how closely the empiricaland the epistemologicalare bound up in
any practicalproblemin historyteaching.The dangeris thatwhileteachersare
on the whole willing to consider empiricalmatters, they often recoil from
philosophicalones, or even try to treat them as if they were empirical.
At the same time there are teacherswho try to follow back the arguments
(arisingas practicaldisputesin the first instance)whereverthey may lead. I
have alreadyreferredto the tendency(for example)of those who advocatea
"skills"approachto minimizethe possibilityof achievinghistoricaltruth, or
even of makingtrue statements.The attractionof this kind of view is that it
appearsto licensechildren'smethodologicalexercisesregardlessof the conclu-
sionsthey reach.It is frequentlysupportedby argumentsstressingthe impossi-
bilityin historyof any directinspectionof the facts to be explained,and some-
times ends in a constructionistposition not unlike that of Goldstein.60There
seemsto be a fatal flawin the assumptionthat because"therealpast"playsno
role in the verificationof historicalassertions,it can be dispensedwith. But no
doubt many of the assumptionsof "traditional"historylead back to equally
disputablepremises.Many examinationpapersappearto manifesta concep-
tion of history in which "the real past"plays an active methodologicalrole.
Nor do Mink'sdoubtsaboutthe occurrenceof "detachableconclusions"in his-
torians'work show any sign of havingfilteredthroughto many of the advo-
cates of "traditional"history.61
Claimsas to what childrencan do in history,are, on the face of it, empiri-
cal; but they too involve sortingout what there is to do in history, and once
again the questions raised are ultimately philosophical. Research into

60. L. J. Goldstein, Historical Knowing (Austin, Texas, 1976). See also History and Theory
Beiheft 16 (1977), "The Constitution of the Historical Past."
61. Louis 0. Mink, "The Autonomy of Historical Understanding," History and Theory 5
(1966), 24-47.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
48 P. J. LEE

children'sthinkingwhichtreatshistoryas if it werethe sameas naturalscience,

or reports about children'simaginativecapacities or empatheticresponses
whichfail to get clearaboutwhatis to count as imaginationor empathyin his-
tory mightfairlybe seen as fallingwithinthe scope of Wittgenstein'sjibe: "ex-
perimentalmethodsand conceptualconfusion."If historicalimaginationwere
just a matter of free association, having images unconstrainedby evidence,
and giving free rein to partialfeelings,then it would not be surprisingto find
that very young childrenare able to displayit.
None of this is to claimthat philosophyof historyon its own can prescribea
correctapproachto the problemsof teachinghistory.Few issuesin philosophy
of historycan be held to have been decided,and it is at least conceivablethat
many of them are essentiallycontested. There is no question of applying
ready-madeanswersto the outstandingproblemsof teachinghistory;philo-
sophicalproblemsare just not "settled"in the requisiteway. And even if con-
clusive answers were available, they would have to be handled with care.
Learninghistoryis not the sameas practicinghistory,andunderstandingis not
an all-or-nothingmatter.Nonetheless,thereis no meritin repeatingclaimsor
marshallingargumentswhich have- at least in the profferedform- already
been shownto be defective;and if matureunderstandingdoes not come all at
once, it needsto be brokeninto digestiblecomponentsbearingsome logicalre-
lationto eachother. Philosophyof historyis necessaryin any attemptto arrive
at a rationalway of teachinghistory, even if it is not sufficient.
The task has only just begun. The particularlogical featureswhich distin-
guish historyfrom other forms of knowledge-obvious candidatesare its nar-
rativestructure,colligatoryorganization,its concernwith humanaction and
rationalexplanation-will obviouslyrequiremuchmore carefulconsideration
in any attemptto characterizethe discipline'smost importantfeaturesas a
basis for historyteaching.Hithertodiscussionhas for the most part concen-
tratedon the notion of historicalevidence;therehas been someworkon cause,
explanation,and empathy, but imagination,interpretation,objectivity,sig-
nificance,and importanceall cry out for more attention.
Any sharpeningof our understandingof these areas offers the possibility
both of greaterclarityin our conceptionof whatit is that childrenshouldlearn
in learninghistory,and also of systematicallybreakingdown the learningand
thinkingtaskswhichchildrenface. It will then be possibleto pick out moreac-
curatelythe difficultiesconfrontingpupils, to simplify more coherentlyand
less misleadingly,and to tracethe relationshipsbetweenthe differentconcepts
and procedures(partly) constitutive of history. Here interest in children's
thinkingand questionsabout the basic structureof the disciplinemerge:but
this mergermust not be at the expenseof seriousconsiderationof criteriafor
selectingsubstantivecontent. If historyhas its own criteriaof significanceand
importance,duckingthese in favor of concentrationon concepts and proce-
duresis in the end to duck a crucialrangeof the conceptsand proceduresin
question.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions
 HISTORY TEACHING AND PHILOSOPHY OF HISTORY 49
The similarityof all this to Bruner'ssketchof a programof honest simplifi-
cationin The Process of Education will be obvious.But Brunerlargelyignored
history,as historiansand historyteachershavelargelyignoredthe philosophi-
cal underpinningsof the argumentsthey seek to conduct.62Thereis alreadya
vast amountof work in philosophyof historywhichthose concernedwith his-
tory teachingmust reckonwith, and alongsidethat moregeneralargumentsin
epistemology,philosophy of mind and action, and (one is tempted to add
"even")philosophyof science. What is lackingstill is any sustainedeffortby
those concernedwith historyin educationto drawon those insights.Perhaps
the trafficneed not be all one-way,and in the end demandsfor a clearercon-
ception of historyfrom those who wish to teach it may raisenew and sharper
questionsin philosophyof history.

Institute of Education,
University of London

62. Bruner,Process of Education. Rogersand Shemiltare conspicuousexceptionsto this rule

amongthose currentlywritingon historyin education.

This content downloaded from 62.122.78.62 on Sun, 22 Jun 2014 08:37:50 AM

All use subject to JSTOR Terms and Conditions