When objects lose their meaning: What happens to their use?

Copyright 2002 Psychonomic Society Inc 236

Cognitive Affective amp Behavioral Neurosciences2002 2 (3) 236-251

Patients with semantic dementia (SD) form the tempo-ral variant of frontotemporal dementia (Hodges PattersonOxbury amp Funnell 1992 Snowden Goulding amp Neary1989) The syndrome is associated with circumscribedtemporal lobe atrophy most severely affecting the tem-poral pole anteromedial and inferolateral temporal lobestypically bilaterally but asymmetrically The ventrome-dial frontal cortex (bilaterally) and the amygdaloid com-plex are also often affected (Mummery et al 2000Mummery Patterson Wise Price amp Hodges 1999) Theselective nature of the semantic deficit in these patientshas been confirmed by their good performance on assess-ments of current day-to-day memory short-term verbalmemory visuospatial skills nonverbal reasoning phonol-ogy and syntax until very late in the course of the disease(Hodges et al 1992 Warrington 1975) More recentstudies have highlighted the patientsrsquo progressive loss ofknowledge about the meanings of nonverbal as well asverbal stimuli including knowledge of objects (BozeatLambon Ralph Garrard Patterson amp Hodges 2000

Hodges Bozeat Lambon Ralph Patterson amp Spatt 2000Lambon Ralph amp Howard 2000)

The syndrome of SD provides an ideal testing groundfor investigating the role of conceptual knowledge in var-ious cognitive domains These include the issue of whetherthe ability to use familiar objects is reliant in whole or inpart upon intact conceptual knowledge about those itemsThere have been reports of dissociations between concep-tual knowledge and object use and this has been taken asevidence for the existence of a separate action semantic sys-tem in which knowledge about the use of objects is storedindependently of other forms of knowledge This evidencewill be evaluated in the Discussion section

In the few studies thus far of object use in SD the ma-jority of patients have shown object use that is impairedto an extent consonant with their conceptual knowledgefor the same items Although these reports highlight thecritical and central role of conceptual knowledge thereare a number of other factors that might be implicated inobject use In the recent study by Hodges et al (2000)for example there were a small number of instances inwhich the patients appeared to use an object more suc-cessfully than would be predicted by their knowledgeabout it We suggested that several other factors might beinfluentialmdashnamely residual conceptual knowledge (a

Correspondence concerning this article should be addressed toS Bozeat MRC Cognition and Brain Sciences Unit 15 Chaucer RoadCambridge CB2 2EF England (e-mail sashabozeatmrc-cbucamacuk)

When objects lose their meaningWhat happens to their use

SASHA BOZEATMRC Cognition and Brain Sciences Unit Cambridge England

MATTHEW A LAMBON RALPHUniversity of Manchester Manchester England

KARALYN PATTERSONMRC Cognition and Brain Sciences Unit Cambridge England

and

JOHN R HODGESMRC Cognition and Brain Sciences Unit Cambridge England

and Addenbrookersquos Hospital Cambridge England

The 8 patients involved in this study were impaired on tests assessing knowledge of objects and onthe demonstration of their use The patientsrsquo success in object use was significantly correlated withtheir knowledge about the objects providing further evidence that conceptual knowledge plays a keyrole in object use Having a recipient present improved performance in the moderately impaired pa-tients suggesting that a certain level of conceptual knowledge must remain for the additional infor-mation to be beneficial Although overall accuracy in using the target objects was not related to ourmeasures of affordance the specific aspects of use afforded by the objectsrsquo structures were relativelyimpervious to semantic impairment suggesting a role for affordance information when object-specificknowledge is disrupted The patientsrsquo familiarity with the objects was an important predictor of per-formance Finally despite good performance on tests of mechanical problem solving the patientsshowed very little evidence of employing these skills in their interactions with real objects

OBJECT USE IN SEMANTIC DEMENTIA 237

relative preservation of general conceptual knowledgesuch as the superordinate category) familiarity personallyrelevant schemata (personal familiarity) and contextsobject affordances and mechanical problem-solvingstrategies As well as having the goal of confirming thatuse is directly tied to the degree of semantic impairmentfor the target object the present study was specificallydesigned to address the role of some of these additionalfactors

Object AffordanceIn 1977 Gibson proposed the theory of affordances in

which he stated that information available more or less di-rectly from the perception of an object gives clues as to itsfunction and possible manipulations (Gibson 1977) Thesearch for the origins of the concept of affordance howevertakes us back to the 1930s and the work of Gestalt psychol-ogist Koffka (1935) who proposed that the meaning orvalue of an object was perceived as readily as its physicalfeatures The term valence was later coined to refer to thisidea There was one crucial difference between the conceptsof affordance and valence Whereas the valence of an ob-ject was thought to be bestowed upon it by the need of theobserver an affordance does not change with the need ofthe observer it is always there to be perceived FollowingGibson the notion of affordance was refined by Neisser(1994) who proposed that the term should be limited tothe notion of physical affordances and not be in any waydependent on stored knowledge about objects In recentstudies the term affordance has been used to refer to twopotentially separate mechanisms that support object usedirectly from visual andor tactile input One is problemsolving or reasoning about the use of an object on the basisof its physical characteristics which will be discussedbelow The other is more like the Gibsonian notion of affor-dance in which clues to the hold orientation movementand purpose of an object are systematically related to itsphysical structure (eg if it has a sharp edge it is used forcutting)

In neuropsychological contexts affordance has typicallybeen characterized in a post hoc fashion If a patientrsquos useof an object is better than would be expected it is assumedthat the physical properties are guiding the patient towardthe correct manipulation For example Hodges et al(2000) reported that the SD patients did use two objects(a pair of scissors and a watering can) better than would beexpected on the basis of their scant remaining conceptualknowledge for these items and suggested that the physi-cal properties of the objects were influencing their useThe construction of a pair of scissors for example is suchthat it is almost impossible not to demonstrate the correctcutting movement In the present study a feature databasewas created not only to improve scoring of object use butalso to identify the systematic relationships between phys-ical features and the way an object is used In this way wehoped to be able to specify affordances a priori before as-sessing their influence on the performance of the SD cases

Mechanical Problem SolvingPatients with SD retain good problem-solving skills

as demonstrated for example by excellent performanceon the Novel Tools test (Goldenberg amp Hagmann 1998)until late in the course of the disease (Hodges et al 2000Hodges Spatt amp Patterson 1999) It is perhaps surpris-ing then that Hodges et al (2000) observed little evi-dence of trial-and-error explorations by the SD patientsthat might have helped them to work out the correct useof real objects The authors proposed two additional con-siderations that may explain this finding First in orderto solve a problem effectively it is necessary to know thetarget or goal The goal is provided by the examiner intests of mechanical problem solving whereas in thedemonstration of single-object use the patients must con-ceive the goal for themselves The ability to deduce thecorrect goal for an object may be dependent on conceptualknowledge (ie the patient must know what the objectrsquoscanonical function is) Second the ability to work out aplausible function for an object often requires knowledgeabout its properties (eg knowledge about the materialsit is made from) which may also be impaired in thesepatients A further possible factor is that the mechanical-problemndashsolving test used in the previous study the NovelTool test (Goldenberg amp Hagmann 1998) can be accom-plished purely through visual matching of the tools andthe blocks making it a nonstringent measure of real prob-lem solving

Several aspects of the present study were designed inorder to explore these possible explanations for the appar-ent absence of problem-solving behavior in real objectuse The patients were assessed on single-object use withand without a recipient (see below) to investigate whetherthe presence of the recipient provides clues as to the goal(the objectrsquos function) thereby enabling better problemsolving We also included a more difficult assessment ofmechanical problem solving mechanical puzzles (basedon those designed by Ochipa Rothi amp Heilman 1992)which cannot be accomplished simply on the basis of vi-sual matching

Presence of a RecipientPrevious studies by our group (Hodges et al 2000

Hodges et al 1999) have assessed use of single objects inisolation to enable investigation of the role of conceptualknowledge without contaminationby other factors that areinvolved in object use Obviously this does not accuratelyreflect our everyday interaction with objects where wewould typically use two or more objects together to carryout a task Indeed this may go some way toward explain-ing how the SD patient reported by Lauro-Grotto Picciniand Shallice (1997) was able to prepare and cook a varietyof different foods despite performing poorly on tests assess-ing conceptual knowledge of the same items We proposethat the recipient may benefit the patients in two waysFirst it may provide a level of context and therefore helpto constrain the conceptual knowledge that underpins cor-

238 BOZEAT LAMBON RALPH PATTERSON AND HODGES

rect object use second as has already been noted it mayprovide a goal for the use of the object and therefore en-courage problem solving

FamiliarityPremorbid familiarity with a concept is an important

predictor of many aspects of performance in SD (Bozeatet al 2000 Funnell 1995 Lambon Ralph Graham Ellisamp Hodges 1998) Previous assessments of object use havetypically involved a corpus of highly familiar objects In thispresent study however items were selected to cover a rangeof familiarity in order to explore its influence on object use

Personally Relevant Schemata and ContextsAlthough this factor is not explored in the present study

it should be noted that repeated experience with personallyfamiliar objects seems to help to maintain appropriate re-sponses to them in the face of severe degradation of con-ceptual knowledge From clinical reports it is clear that pa-tients with SD manage well with everyday tasks and oftencontinue with hobbies until late in the course of the disease(Graham Lambon Ralph amp Hodges 1997 Lauro-Grottoet al 1997 Snowden Griff iths amp Neary 1994) Two re-cent studies addressed this issue directly (Bozeat Lam-ben Ralph Patterson amp Hodges 2002 Funnell 2001)In both it was found that SD patients were more accuratewhen demonstrating the use of their own objects thanwith equally good substitute exemplars

METHOD

PatientsEight patients were identified through the Memory and Cogni-

tive Disorders Clinic at Addenbrookersquos Hospital Cambridge Eng-land where they were seen by a senior neurologist (JRH) a seniorpsychiatrist and a clinical neuropsychologist In addition to a clin-ical assessment all the patients were given a number of standardpsychiatric rating scales in order to exclude major functional psy-chiatric disorders such as depression and schizophrenia They all un-derwent MRI scanning together with the usual battery of screeningblood tests in order to exclude treatable causes of dementia

All the patients presented with a progressive loss of vocabularythat affected expressive and receptive language in the context of flu-ent speech production They all fulfilled previously proposed crite-ria for SD anomia impairment in single-word comprehension andimpoverished semantic knowledge with relative preservation ofphonology syntax visuospatial abilities and day-to-day memory(Hodges Graham amp Patterson 1995 Hodges et al 1992) In all 8

cases structural brain imaging by MRI showed focal atrophy in-volving the polar and inferolateral regions of the temporal lobeson the left side only for the mildest case (AN) but bilateral in allof the others (see Table 1) In 6 of the 7 bilateral cases there wasmore extensive atrophy in the left than in the right temporal lobe(L R in Table 1) KH being the one exception (R L) In themajority of cases (except for the 3 milder onesmdashAN JC andAT) the atrophy had spread from the temporal pole anteriorly tothe ventromedial frontal region as is typical in SD (Mummeryet al 2000 Snowden Neary amp Mann 1996)

Ten normal subjects from the Medical Research Council Cognitionand Brain Sciences Unitrsquos subjectsrsquo panel approximately matchedin age and education to the patients served as control subjects

General NeuropsychologyThe following battery of neuropsychological tests was adminis-

tered the Mini-Mental State Examination as a general measure ofcognitive impairment (Folstein Folstein amp McHugh 1975) the digitspan subtest of the Wechsler Memory ScalendashRevised (WMSndashRWechsler 1981) to assess auditory-verbal short-term memory ver-bal fluency for the letters F A and S to test executive function theRavenrsquos Colored Progressive Matrices to assess nonverbal problemsolving (Raven 1962) copy and immediate recall of the Rey Com-plex Figure to test visuospatial skills and episodic memory (Rey1941) Various subtests from the Visual Object and Space Percep-tion battery were also used to assess visuospatial function in moredetail (Warrington amp James 1986)

Semantic AssessmentsThe patients were given a selection of tasks from a semantic bat-

tery which is a collection of tests that use the same set of stimulusitems to assess semantic knowledge systematically across differentinput and output modalities It contains 64 items representing threecategories of living things (animals birds and fruit) and three cate-gories of artifacts (household items tools and vehicles) The follow-ing subtests from the semantic battery were administered categoryfluency in which the subject is asked to produce as many exemplarsas possible in 1 min for each of the six categories naming of the 64line drawings spoken word-to-picture matching using picture arrayscontaining the target plus nine within-category foils In addition thePyramid and Palm Trees test of associative semantic knowledge(Howard amp Patterson 1992) was also administered In this assess-ment the subjects are asked to choose one of two items that is mostclosely associated with the target (eg for the target pyramid thechoice is between palm tree and pine tree) The stimuli are pre-sented as either pictures or written words

General Praxis TestingThe subjects were asked to copy the examiner in producing 10

meaningless gestures (Goldenberg 1996) Two points were awardedfor a correct gesture on the first attempt and a single point if it wascorrect on the second presentation

Table 1Demographic and Imaging Data

Control

AN JC AT KH DS JH DC BW M SD

Age 62 59 65 58 61 59 78 70 637 62

Sex M M M M M F F F 55MRI Mild left Moderate Moderate Moderate Moderate Marked Marked Marked

temporal bilateral bilateral bilateral bilateral bilateral bilateral bilateralatrophy temporal temporal frontotemporal frontotemporal frontotemporal frontotemporal frontotemporal

atrophy L R atrophy L R atrophy R L atrophy L R L R L R L R

Ratio of male to female


Mechanical Problem Solving 1 The Novel Tools test (Goldenberg amp Hagmann 1998) The

materials for this test consist of six wooden cylinders each of whichcan sit in a wooden base and a selection of novel tools Each cylin-der has a part to which one of the tools can be fitted to lift the cylin-der out of its base During testing one cylinder at a time is placedin the well of the base and a collection of three tools is placed be-side it The subject is asked to select the tool best suited to lift outthe cylinder If the correct tool is not chosen as the initial responsethe subject is asked to choose an alternative

Two aspects of this task were scored separately the selection andthe use of the correct tool For the first part two points were givenwhen the correct tool was selected at first choice and one point wasgiven if the subject selected the correct tool on the second choice(maximum score = 12) The second part of the test evaluated the useof the tool (either selected by the subject or given by the examinerfollowing two incorrect selections) Two points were awarded if thesubject inserted the tool and lifted the cylinder without hesitation orerror and one point was awarded if the subject demonstrated thecorrect use after trial and error (maximum score = 12)

2 Mechanical Puzzles (Ochipa et al 1992) The stimuli consistedof nine clear Perspex cylinders each containing a wooden blockand a selection of novel tools The goal was always to remove theblock from the cylinder and each task required both the selectionof the appropriate tool and the use of a different procedure strategy(eg lifting levering pushing pulling etc) During testing one cylin-der at a time was placed in front of the subject and a collection offour tools was placed beside it The subject was asked to select thetool best suited to remove the block out of the cylinder

Once again the two aspects of this task were scored separatelyFor the first part one point was given when the correct tool was se-lected at first choice (maximum score = 9) To evaluate use of thetool (either selected by the subject or given by the examiner follow-ing an incorrect selection) two points were awarded if the subjectinserted the tool and removed the block without hesitation or errorand one point was awarded if the subject demonstrated the correctuse after trial and error (maximum score = 18)

Object Use BatteryA multiple component battery was constructed with the purpose of

assessing associative information functional knowledge and use of36 household objects These were derived from three categoriesmdashtools kitchen implements and stationery itemsmdashand covered a fairlywide range of rated familiarity

Tests of conceptual knowledge Conceptual knowledge for the 36objects was assessed in a series of matching tests which consistedof digital photographs of the targets and similar photographs of fourpossible matches for each target The picture of the target object waslocated at the top of the page and the subject was asked to chooseone of the four response alternatives as the best match according toone of three types of relationship described below The order of itemswas randomized across tasks and each was preceded by four prac-tice trials

Every effort was made to ensure comprehension of the task Datawere not included if there was any doubt about the patientsrsquo abilityto comprehend the instructions which occurred in 3 subjects (JHDC and BW) when they were asked to match according to the ac-tion that would be used when the target object was manipulated

1A Matching to Recipient The subjects were asked to choose thecorrect typical recipient for the target object The foils were chosento be visually similar to the correct match or semantically related(eg for the target garlic press the recipient choice is between gar-lic onion pepper and cheese)

1B Matching to Function In this test the subjects were asked tochoose one of four objects that could be used for the same purposeas the target item The foils were chosen to be either visually simi-lar to or from the same category as the target (eg for the target gar-

lic press the choice is between pestle and mortar corkscrew scis-sors and pliers)

1C Matching to Action In this test the subjects were asked tochoose one of four objects that would be manipulated moved in thesame way as the target The object that represented the correctchoice is not necessarily held in the same way as the target but re-quires a similar action The foils were chosen to be visually similaror semantically related to the target (eg for the target garlic pressthe choice is between secateurs corkscrew bottle opener and com-passes)

2 Naming The subjects were given each object individually andwere asked to produce the name

3 Word-to-picture matching The subject was asked to choosethe item from picture arrays containing the target plus seven within-category foils in response to the spoken target name

4 Action-to-picture matching In this test use of the object wasmimed by the experimenter and the subject was asked to choose theobject being used from an array of eight within-category itemsThese were the same arrays as those used in the word-to-picturematching test Data were not included for one patient DC whowas unable to comprehend this task

5A Single-object use The subjects were given each real objectin isolation and were asked to demonstrate its use Performance wasvideotaped for later evaluation

5B Object use with recipient Only a subset of the 36 items weretested with a recipient (n = 22) because some were not practical(eg potato masher) and others did not have a typical recipient(eg tape measure) The subjects were given each object and its re-cipient and were asked to demonstrate their use together Perfor-mance was videotaped for later evaluation

Scoring Object UseA feature database was constructed to enable the quantification of

the physical affordances of the objects (details below) and also tocreate a feature-based scoring scheme in place of the rater-basedmethods used in many other studies of object use Object use by thecontrol subjects was examined first and from these data a templatewas formed for each object This template specified a description ofthe canonical use of each object in terms of composite features thenumber of hands used to hold the object the grasp the position onthe object and each individual movement So for example the fea-tural description for the use of a hammer was the following held inone hand with a ldquostandardrdquo grasp about half-way down the handlewith the flat end of the head facing down lift-up strike down move-ment repeated The control subjectsrsquo and the patientsrsquo object usevideos were then scored according to this template leading to sep-arate scores for the hold movement and orientation of the objectBecause the total possible score for the hold and movement variedacross objects proportional scores were used in all the analyses

RESULTS

General NeuropsychologyThe 8 patients covered a broad spectrum of impair-

ment as is indicated by their performance on the Mini-Mental State Examination (see Table 2 in which the pa-tients are ordered by their overall performance on the threeobject-matching tasks) Five of the 8 cases showed intactworking memory as measured by forward and backwarddigit span DCrsquos forward span was (just) within normallimits but she could not comprehend the instructions forbackward span DS and BW who like DC had a pro-found anomic aphasia were also slightly subnormal evenon forward verbal span All the patients except AN ex-


hibited some impairment on the letter fluency test with thethree most anomic cases being outliers There was generalpreservation of nonverbal problem-solving and visuo-spatial skills as measured by Ravenrsquos Colored ProgressiveMatrices and the Visual Object and Space Perceptionbattery

Semantic AssessmentsAs is shown in Table 3 the patients included in this

study covered a wide range of semantic decline from thevery mildly impaired patient AN whose deficits wereonly revealed by a subset of the more taxing assessmentsto BW who had profound semantic degradation All thepatients except AN showed reduced category fluencysome degree of anomia as indicated by their performanceon the naming test and impaired comprehension as mea-sured by the word-to-picture matching (note that the pa-tients in Tables 2 and 3 are ordered as best as possible to re-flect decreasing scores on the three object-matching tests)A number of patients with semantic deficits especiallyfrom herpes simplex virus encephalitis have been reportedin the literature to show dissociations in performance ac-cording to category of knowledge (eg living vs man-made) Looking at the performance on the word-to-picturendashmatching test all but one of the impaired patients had nu-

merically better scores on items from the artifact domainalthough this difference was very small in most casesOne patient DS performed considerably better on themanmade items (2532 vs 1532) but it should be notedthat his performance on items from both domains is wellbelow that of the control subjects All the patients were im-paired on both conditions of the Pyramid and Palm Treestest except for AN whose performance was normal onthe picture version

General Praxis TestingAn independent samples t test revealed no difference

between the performances of the patients and the controlsubjects on copying the meaningless gestures [t(16) =157 ns] indicating that the patients suffered from nosignificant impairments to general praxis

Mechanical Problem SolvingThe patients performed well on both selection and use

of the tools in both mechanical-problemndashsolving tasks(see Figure 1) A repeated measures analysis of variance(ANOVA) confirmed that there was no reliable differencebetween the performances of the patients and the controls[F(116) = 314 p 05] There was a significant effectof task component [F(348) = 1989 p 001] with se-

Table 3Assessment of Semantics

Subject Control

Test (Maximum Score) AN JC AT KH DS JH DC BW M SD

Category fluencyLiving 47 17 14 8 0 5 0 0 603 126Manmade 34 23 18 14 0 7 0 0 548 103

Naming (64) 64 41 17 42 1 6 2 1 623 16Wordndashpicture matching (64) 64 56 57 51 40 18 23 8 637 05

Living 32 25 27 23 15 10 9 3 318 04Manmade 32 31 30 28 25 8 14 5 320 02

Pyramids and Palm TreesWords (52) 48 44 45 40 40 25 NT NT 511 11Pictures (52) 51 41 47 42 41 34 29 30 512 14

NotemdashNT not tested

Table 2General Neuropsychology

Subject Control


MMSE (30) 30 24 25 22 12 7 8 7 288 05Digit span

Forward 7 7 8 6 4 6 5 4 68 09Backward 7 4 5 5 4 5 0 3 47 12

Letter fluency (total FAS) 40 22 20 13 2 8 0 0 442 112Ravenrsquos colored matrices 95 75 90ndash95 95 75ndash90 95 95 75VOSP

Incomplete letters (20) 20 18 20 20 20 17 NT 19 192 08Dot counting (10) 10 9 10 10 10 10 10 10 99 03Position discrimination (20) 20 19 19 20 20 19 NT NT 198 06Cube analysis (10) 10 10 10 10 10 10 10 9 97 25Number location (10) 10 9 10 10 10 10 NT NT 89 28

NotemdashPatients are ordered according to their performance on the object matching tests VOSP Visual Object andSpace Perception battery NT not tested Scores on the Ravenrsquos matrices are given as percentiles


lection (not surprisingly) proving more difficult than usebut no interaction between group and task component[F(348) 1]

Object Use BatteryObject matching tests Every patient was impaired on

all three matching subtests except for AN who per-formed within the normal range on one of the threematching to recipient (see Figure 2 in which the patientsare ordered by their performance on these three tasks)The three most impaired patients do not have scores forthe matching-to-action subtest because they were unableto comprehend the instructions for this component A re-peated measures ANOVA revealed a significant main ef-fect of group [controls patients F(113) = 522 p 001] and test [F(226) = 109 p 001] but no inter-

action between these two factors [F(339) = 15 p 05]Post hoc tests confirmed that the patientsrsquo scores on allthree matching tests were significantly lower than thescores of the control subjects (t values between 49 and72 all ps 01) Numerically speaking the patients per-formed best on matching to recipient and most poorly onmatching to function Perhaps owing to the high variabil-ity within the group none of the differences betweenthese three tests was statistically significant (t values be-tween 048 and 108 ns) In terms of a different criterion(the ability to comprehend and therefore complete thetask) performance was worst on matching to action

The control subjects performed better on matching torecipient than on either of the other two matching tests(t values of 27 and 45 both ps 05) and their scoreson matching to function were significantly higher than

Figure 1 Performance on the Novel Tools test and mechanical puzzles

Figure 2 Performance on the three tests of visual associative knowledge Data arenot included for three patients JH DC and BW on the action-matching test be-cause they were unable to comprehend the task


their scores on matching to action [t(9) = 34 p 01]Sirigu Duhamel and Poncet (1991) reported a patient

who was unable to recognize the function of objects butwas nevertheless often able to demonstrate appropriatemanipulations(this case will be considered in further detailin the Discussion section) The opposite dissociation pre-served function knowledge and impaired manipulationknowledge was reported in two cases by Buxbaum Ver-amonti and Schwartz (2000) In the present data therewas no significant difference between the performance ofthe patients as a group on matching to function versusmatching to action although the performance of the con-trol subjects indicated that the action task was more dif-ficult As individuals (see Figure 2) 2 of the patients(DS and KH) showed a reversal of the relative difficultyin these two subtests relative to the control subjects (iebetter performance on matching to action than on match-ing to function) but this difference was not significant ineither case (both c 2 1 ns)

Naming The patients were impaired all except ANprofoundly so at naming the objects (see Figure 3) Fourpatients (DS DC JH and BW) failed to name any ofthe 36 objects A t test confirmed that the patientsrsquo scoreson naming were significantly lower than those of the con-trol subjects [t(16) = 938 p 001]

Word-to-picture matching and action-to-picturematching It is clear from Figure 4 that all the patientswere impaired at selecting the objects in response to boththeir spoken names and a pantomime of their use It wasconfirmed with t tests that the patientsrsquo scores on boththese tasks were significantly lower than the scores of thecontrol subjectsrsquo [t(16) = 492 p 01 t(15) = 66 p 001] There was no difference between the patientsrsquo per-formances on these two tasks [t(6) 1]

Object UseFigure 5 reveals that the patientsrsquo ability to demonstrate

the correct use of the objects was poorer than that of thecontrol subjects on all three dimensions (hold move-ment and orientation) A repeated measures ANOVA re-vealed significant main effects of group [F(116) = 341p 001] and object use component [F(232) = 311p 001] plus an interaction between these two factors[F(232) = 141 p 001] The scores of the patientswere significantly higher on correct hold for the objectsthan on either movement [t(7) = 568 p 05] or orien-tation [t(7) = 264 p 05] and were significantlyhigher on orientation than on movement [t(7) = 353 p 05] The control subjects scored best on the orientationcomponent with scores on this component being slightlybut nevertheless significantly higher than scores on thehold [t(9) = 251 p 05] and the movement of the ob-jects [t(9) = 506 p 001] Like the patients the controlsubjectsrsquo scores on correct hold were significantly higherthan those on movement [t(9) = 601 p 001]

The Relationship BetweenConceptual Knowledge and Object Use

Pearsonrsquos correlations revealed significant associationsbetween virtually all combinations of the patientsrsquo scoreson the various semantic tests the five from the semanticbattery (category fluency picture naming word-to-picturematching and the word and picture conditions of thePyramid and Palm Trees test) and the two designed forthis study (naming the objects and a combined score forthe three associative matching tasks 63 r 95 allpsone-tailed 05) The only exceptions were the correla-tions between the word condition of the Pyramid and PalmTrees test and two naming tests which failed to reach con-

Figure 3 Performance on naming the objects


ventional levels of significance (r = 62 p = 09 r = 61p = 010) perhaps because 2 of the patients were not testedon the word condition of the Pyramid and Palm Trees testThese correlations support the view that the impairmentsin these patients reflect damage to a central amodal sys-tem that underpins conceptual knowledge (Bozeat et al2000 Lambon Ralph amp Howard 2000)

One of the aims of this study was to replicate the resultsof the previous study reported by Hodges et al (2000)mdashthat is to demonstrate the importance of conceptual knowl-edge in object use In keeping with this hypothesis bothoverall use and each of the individual components of ob-ject use (hold movement and orientation) correlated re-

liably with all of the semantic tests designed for this studyand with virtually all of the semantic assessments re-ported in Table 3 (68 r 91 all psone-tailed 05)Only correlations of the word condition of the Pyramidand Palm Trees test with movement and overall usefailed to reach conventional levels of significance (r =64 p = 08 r = 67 p = 07)

By-subjects regression analyses were carried out to de-termine whether any individual patientrsquos performance wasdiscrepant from the significant group-based relationshipbetween object use (the total score on the three compo-nents) and knowledge (as measured by the total score onthe three associative matching tests and word-to-picture

Figure 4 Performance on the word-to-picture matching (WPM) and action-to-picture matching (APM) tests Data are not included for DC on APM because shewas unable to comprehend the task

Figure 5 Performance on the individual components of object use


matching) With two standard residuals either side set asthe criterion none of the 8 patients deviated significantlyfrom this relationship

AffordancesSpecifying affordances empirically Gibsonrsquos theory

of affordances proposes that information available directlyfrom perception gives clues as to the function of an objectand the possible manipulations of it (Gibson 1977) Afeature database was constructed to enable a priori quan-tification of these affordances The database contained90 manmade objects and each one was rated accordingto a large number of structural features (n = 56) includ-ing overall size the number of handles the type of han-dle(s) the position of the handle in relation to the end ofthe tool the presence of moving parts and what was atthe end of the tool Various features of the hold (n = 11eg the number of hands position on the tool grasp) ofeach individual movement (n = 17 eg lift up strikedown) and of the function of each object (n = 21 eg cut-ting cleaning) were also specified Systematic relation-ships between features were highlighted by computingSpearmanrsquos correlations for each possible pairing acrossthe 90 objects In total there were 105 features which ledto 5460 possible featurendashfeature pairings it was surpris-ing therefore to find only 46 significant correlations

It is important to consider the chance level when per-forming such a large number of correlations In this caseone would expect 273 significant correlations to occurby chance The number of observed reliable correlationswas therefore significantly lower than would be ex-pected by chance (z = 21410 p 001)

The significant correlations obtained can be classifiedin the following ways (1) structural featurendashstructuralfeature (n = 11 eg if the object has two handles it is

likely to have moving parts) (2) structural featurendashhold(n = 8 eg if the object has a handle that joins the shaftit is likely to be held in a ldquostandardrdquo grasp) (3) structuralfeaturendashmovement (n = 5 eg if the object has a wedge-shaped head it is likely to be associated with a strikingdown movement) (4) structural featurendashfunction (n = 11eg if the object has a sharp serrated edge it is likely tobe used for cutting) (5) holdndashhold (n = 2 eg if one handis a ldquopinchrdquo grip the other hand is likely to be a ldquopinchrdquogrip as well) (6) holdndashmovement (n = 1 eg if the sec-ond hand is a ldquopinchrdquo grip it is likely to be twisted hor-izontally with the fingers) (7) movementndashmovement(n = 5 eg if the object is ldquolifted-uprdquo it is likely to be as-sociated with a ldquostriking-downrdquo movement as well )(8) movementndashfunction (n = 3 eg if the object is heldstill it is likely to be used for measuring)

Do affordances influence object use Twelve affor-danced objects were selected on the basis that either thehold or the movement was reliably predicted by a structuralcharacteristic in the analysis of the feature database de-scribed above These were matched on the basis of famil-iarity to 12 other objects for which neither the hold nor themovement was obviously afforded by their structure Per-formances on the affordanced and unaffordanced set werecompared both for overall use score and on the particulartarget component of use (hold or movement) No differencewas revealed by t tests on scores of overall use [t(11) 1]or of the particular component that was afforded [t(11) =158 p 05] It is clear from Figure 6 however that ob-ject use by some of the patients benefited from these af-fordances and the level of semantic impairment appears tobe an important factor When the patients were subdividedinto two groups according to their level of semantic impair-ment the more impaired patients (n = 4) achieved signifi-cantly better performance on the particular component of

Figure 6 Performance on affordanced and unaffordanced objects In some objects thehold is afforded whereas in others it is the movement In all cases however the score de-picted is for overall use


use that was afforded as compared with objects with no suchspecific affordances [F(13) = 136 p 05] whereas themildly impaired patients showed no difference [F(13) 1n = 4] In the most impaired patients (JH DC and BW)there was also an interaction between affordance and famil-iarity [F(12) = 326 p 05] suggesting that familiarityis only important in the use of unaffordanced items for af-fordanced objects there was no influence of familiarity

The difference between component use scores on theaffordanced and unaffordanced items correlated signifi-cantly with overall semantic knowledge scores (r = 286p 01) This demonstrates in a different way that the levelof semantic impairment is a critical factor in determiningthe impact of affordances on object use

Presence of RecipientOur everyday interaction with objects typically in-

volves using pairs of objects together (one object and its re-cipient) to complete a task (eg using a hammer to drivea nail a corkscrew to open a bottle of wine a potato masherto mash potatoes etc) As was explained in the Methodsection in order to explore the impact of the recipient wereassessed use of 22 of the objects on a different occa-sion with the recipient present

Five of the patients showed a numerical advantage foroverall object use with the recipient present This differ-ence was very small in the 2 patients with mild semanticimpairment (AN and AT) but was quite striking in 3 pa-tients with more moderate semantic impairment (see Pa-tients JC DS and KH in Figure 7) Analysis of the pa-tients as a group revealed that the scores on correct holdfor the objects were significantly higher when the recip-ient was also present [F(17) = 104 p 05] scores werealso higher on correct movement although this did notreach significance [F(17) = 45 p = 07] There was no

difference between these two conditions on scores of ori-entation [F(17) 1] or overall use [F(17) 1] The pa-tients were then subdivided into three groups accordingto their level of semantic impairment a repeated mea-sures ANOVA revealed significant effects of severitygroup [F(25) = 166 p 01] and presence of recipient[F(15) = 1738 p 001] and a significant interactionbetween group and recipient [F(25) = 1639 p 001]Post hoc tests confirmed that only the moderately impairedpatients (n = 3) scored significantly better with the recip-ient present [t(2) = 177 p 01] there was no differ-ence between performance with and without a recipientin the mildest patients [t(1) = 5 p 05 n = 2] or in themost impaired patients [t(2) = 302 p 05 n = 3]

FamiliarityFamiliarity is an important predictor of performance on

tasks assessing conceptual knowledge (Bozeat et al 2000Funnell 1995 Lambon Ralph et al 1998) so it was pre-dicted that it would be an important factor in object useFamiliarity ratings were initially obtained by asking 20normal age-matched subjects to rate how often they useeach object The 36 items in the object use battery werechosen to cover a range from highly familiar items that areused by most people on a daily basis (eg a pencil) to lessfamiliar items that are used by most people only aboutonce a year (eg a chisel ) From inspection of these rat-ings it became clear that familiarity varies greatly fromone person to another being highly dependent on careerand lifestyle These ratings were used to create the bat-tery of items but it was decided that they would not besuitable for analysis of the effects of familiarity on thepatientsrsquo object use Most of the patients involved in thisstudy had been suffering from SD for several years and asa consequence their hobbies and daily activities were

Figure 7 Performance on single-object use and use with recipient


greatly reduced Ratings of familiarity were therefore ob-tained for each patient from his or her spouse or caregiverPearsonrsquos correlations revealed a significant associationbetween the patientsrsquo success in overall object use and thefamiliarity ratings collected from the patient caregivers(r = 39 pone-tailed 01)

Although not the main point of this analysis it is of in-terest to note that the spousecaregiver ratings indicatedthat the patients did indeed have much less contact withmost of the objects than did the control subjects A t testconfirmed that the familiarity ratings for the patients weresignificantly lower than the ratings obtained from thecontrol subjects [t(7) = 1153 p 001] It is furthermoreinteresting to note however that some of the patientswere assigned surprisingly high familiarity ratings withsome objects that from the ratings obtained from the con-trol subjects were deemed to be relatively low in famil-iarity For example DC was reported to use a tape mea-sure every day to measure the length and width of jigsawpuzzle boxes in order to cut pieces of Sellotape to the exactsize for fastening the boxes

Problem SolvingAll the patients performed well on the tests of mechan-

ical problem solving the Novel Tool test and the mechan-

ical puzzles We wanted to determine therefore whetherthey were utilizing these good problem-solving skills intheir use of real objects Because we had predicted that pres-ence of a recipient might enhance problem-solving behav-ior the first analysis compared object use with and with-out a recipient

Overall there was no significant difference in the rateof problem solving (defined by at least two attempts to usean object in different ways) between use of the objectswith and without a recipient [t(8) = 202 p = 08] Only2 patients (KH and JH) were found to use trial anderror consistently across a number of items In order toexplore the impact of this problem-solving behavior inthese two cases we compared object use scores on thefirst attempt with those achieved on the last attempt Theanalysis produced mixed results with KH showing asignificant improvement on one component of object usewhereas JH demonstrated no improvement on any of thecomponents (see Figure 8) The score achieved by KHon the movement of the objects was significantly higherfollowing problem solving [t(13) = 38 p 01] hisoverall use was also better following problem solvingalthough this did not reach significance [t(13) = 19 p =07] There was no difference between scores on hold[t(13) 1] or orientation [t(13) 1] before and after

Figure 8 (A) Performance of Patient KH before and after problemsolving (B) Performance of Patient JH before and after problemsolving


problem solving JH showed no improvement on over-all use [t(12) 1] or on any of the individual compo-nents [hold t(12) = 139 p 05 movement t(12) 1]in fact her scores on orientation were significantly higherbefore trial-and-error problem solving [t(12) = 274 p 05] These results suggest that the patients do not oftenmake use of their good problem-solving skills to workout what to do with objects and even when they do itneed not be beneficial for all aspects of use

DISCUSSION

In a previous study (Hodges et al 2000) competencein the use of familiar objects by patients with SD was sig-nificantly predicted by the patientsrsquo degree of retaineddisrupted conceptual knowledge for the same objects Theprimary aims of the present study were (1) to replicate theresults of the previous study using a more comprehensivebattery of conceptual knowledge tests a larger corpus ofitems and a feature-based approach to scoring object useand (2) to investigate the influence of a range of other fac-tors that may impact on object use including object affor-dance presence of a recipient familiarity and mechanical-problemndashsolving strategies The results for each of thesefactors is summarized and discussed in turn below

Conceptual Knowledge Seven of the 8 patients involved in this study were im-

paired on all the tests assessing conceptual knowledgewith one case (AN) showing deficits on a subset of theseassessments All the patients (again with the exceptionof AN) were impaired at demonstrating the use of theobjects and across the 8 cases success in object use wassignificantly correlated with level of conceptual deficitTaken together with the results of our previous study(Hodges et al 2000) and those of Hamanaka and collegues(Hamanaka et al 1996) 18 cases of SD have now beenreported in which deficits in object use in line with the pa-tientsrsquo conceptual impairment have been shown It is alsoimportant to note that the impact of two other factorsmdashnamely the presence of a recipient and affordancemdashwasmodulated by the level of conceptual impairment (thisfinding will be discussed in further detail below) Theseresults provide strong evidence for the key role played byconceptual knowledge in object use

Dissociations between knowledge about an objectrsquosfunction and its manipulation have been reported in theliterature (Buxbaum et al 2000 Sirigu et al 1991) andit has consequently been suggested that certain types ofconceptual knowledge about objects may be more criticalfor their use than are others We found no evidence fordissociations between different types of knowledge Thepatients were equally impaired on all aspects of conceptualknowledge We should emphasize however that this con-clusion applies to the use of single objects as was assessedhere There may be other forms of knowledge which maybe conceptual or more accurately described as proce-

dural that help to support action in naturalistic settingswhere the patient has (1) a goal in mind and (2) a wholerelevant context in which to act on and with the object(s)

Impaired Object Use in theContext of Preserved Semantic Knowledge

Several patients have been reported in the literature whowere unable to use real objects correctly despite havingpreserved knowledge about those same objects (RumiatiZanini Vorano amp Shallice 2001 Spatt Bak Bozeat Pat-terson amp Hodges 2002) These patients invariably hadsome level of ideomotor apraxia associated with damage toparietal regions which left them unable to produce themovements appropriate for object use There has beensome controversy in the literature as to whether the con-cept of ideomotor apraxia should be limited to tests ofpantomime and imitation or whether it also has an im-pact on real object use Zangwill (1960) noted that diffi-culties in using real objects may be related to a severe pro-duction disorder In concordance with this we havereported a group of patients with ideomotor apraxia owingto corticobasal degenerationwho had difficulties demon-strating the use of real objects (Spatt et al 2002)

Ochipa Rothi and Heilman (1989) reported a left-handed patient who following a right-hemisphere strokewas able to name objects but was unable to point to themwhen their functions were described or to describe theirfunctions himself Furthermore he was unable to demon-strate their uses This inability to use tools could not beexplained solely by a production deficit because he wasalso unable to match tools to their recipients suggestingan impairment in the appreciation of the functional rela-tionship between different objects The authors proposedthat this patient was suffering from an impairment in theaction semantic system Closer inspection of these datahowever suggested deficits on other semantic tasks aswell For example the patient succeeded in naming 1720of the objects in the experimental battery (no control datawere reported but these objects were described as ldquocom-mon household tools and objectsrdquo implying that most peo-ple would perform at ceiling on this task) and he scoredjust 4860 on an alternative naming task His perfor-mance was undoubtedly better on general semantic tasksthan on tasks assessing knowledge of tool use but thispattern of results is perhaps explicable in terms of taskdifficulty Describing the function of objects and indeedselecting objects in response to descriptions of their func-tion are more linguistically demanding tasks than simplynaming objects or selecting them in response to their spo-ken names

Furthermore this patient had suffered fairly extensivebrain damage (including frontal inferior parietal and su-perior temporal regions) which is likely to have affected anumber of cognitive domains Although the authors arguedthat the semantic impairment was confined to the actiondomain it seems plausible that there was at least a degreeof impairment in general semantic knowledge Further-


more the patient was observed to misuse common objectsin their natural settings in a manner suggestive of a frontaldysexecutive syndrome It is therefore not so clear that thispatient had selective damage to an action semantic system

Moreaud Charnallet and Pellat (1998) reported thesame dissociationmdashimpaired object use in the context ofpreserved conceptual knowledgemdashin a patient with mod-erate stage Alzheimerrsquos disease Despite performing wellon tests tapping knowledge of 15 common household ob-jects and preserved praxis this patient was not alwaysable to demonstrate their use correctly Once again how-ever careful inspection of these data revealed that the pa-tient did not always perform well on the tasks assessingconceptual knowledge For example EJ was able to pro-vide names and describe the use of only 3 of the 7 objectsthat he failed to use correctly In fact there were only 2objects that EJ failed to use despite demonstrating pre-served knowledge as assessed by all the semantic tasksThese objects were a camera and a corkscrew which de-pending on the exemplar can be fairly complicated touse Like the case reported by Ochipa et al (1989) thispatient was also reported to show marked difficulties withexecutive functioning

Preserved Object Use in theContext of Degraded Semantic Knowledge

Patients with SD seem to manage surprisingly wellwith everyday tasks and have been reported to use a num-ber of objects correctly even the same objects to whichthey cannot provide names descriptions or correct asso-ciative semantic judgments Such observations are how-ever largely anecdotal with few investigations havingsystematically explored the use of real objects BuxbaumSchwartz and Carew (1997) reported a patient who de-spite a moderate degree of semantic impairment usedmost objects normally In this study however the authorsdid not assess knowledge about and usage of the same ob-jects precluding a definitive conclusion that this patientwas able to use objects for which he had degraded se-mantic knowledge A study by Lauro-Grotto et al (1997)assessed the ability of another patient with SD to preparefood which she did without error for nearly all ingredi-ents despite performing poorly on verbal tests assessingknowledge of the same items This study however didnot assess single-object use and it is possible that the pa-tientrsquos successful use of kitchen tools and ingredientsmay have benefited strongly from the rich contextual en-vironment in which she was tested

In contrast three studies have concluded that semanticimpairment does lead to deficits in object use Hamanakaet al (1996) reported the co-occurrence of impoverishedconceptual knowledge and impaired object use in two SDpatients There is some indication from this report thatthe degree of semantic impairment may be a critical fac-tor One of the patients initially presented with a mild se-mantic deficit affecting verbal comprehension and pro-duction and at that stage had preserved object use Over

time however as the patientrsquos comprehension deterioratedfurther the ability to use common objects declined tooHodges et al (1999) described two SD patients with se-vere loss of conceptual knowledge about objects associ-ated with many failures to use the same items correctly

In a follow-up study we investigated the role of concep-tual knowledge in object use with a comprehensive batteryof tests devised to assess associative information func-tional knowledge and use of 20 common objects (Hodgeset al 2000) In addition to this battery of tests the 9 SDpatients were assessed on measures of general praxis andmechanical problem solving Object use was found to bemarkedly impaired and this could not be explained byproblems with general praxis since the patients performedwell on copying of the meaningless gestures Impor-tantly the patientsrsquo success in demonstrating the use ofobjects correlated strongly with their performance on nam-ing of and semantic knowledge of the same objects Fromthese data we concluded that conceptual knowledge playsa key role in object use

The pattern of deficits seen in patients with optic apha-sia is also often cited as evidence for a dissociation be-tween impaired semantics (or in this case impaired visualaccess to semantics) and preserved knowledge of objectuse These patients have difficulty naming visually pre-sented objects and pictures but can name the same items inresponse to tactile presentation or auditory definitions(Riddoch amp Humphreys 1987) Most striking is the ob-servation that patients with optic aphasia apparently canoften demonstrate the appropriate use by gesture of ob-jects that they fail to name upon visual confrontation Thispattern of performance however does not require an in-terpretation of preserved action semantics RiddochHumphreys Coltheart and Funnell (1988) influencedby the work of direct perceptionists such as Marr andGibson suggested that these gestures were being madeon the basis of nonsemantic forms of information theperceptual attributes of the objects andor appropriate ac-cess to a stage of processing termed structural descrip-tions of objects (Humphreys amp Forde 2000) that is in-termediate between perception and semantics

A case reported by Sirigu et al (1991) further illustratesthe influence of these nonsemantic forms of informationThis associative agnosic patient (FB) had poor knowl-edge of the functional and associative attributes of ob-jects When asked to describe how he would use variousobjects and to demonstrate correct use from sight how-ever his descriptions and manipulations invariably re-spected the mechanical affordances of the object but notnecessarily its conventional function For instance forthe iron he said ldquoyou hold it one hand and move it backand forth horizontally [miming the action] Maybe you canspread glue evenly with itrdquo Sirigu et al argued that FBwas able to achieve a precise analysis of the mechanicalproperties of the objects and that visual and or tactile in-puts were able to trigger sensory motor representationswhich in turn permitted appropriate action independent of


the semantic system It is important to emphasize how-ever that even though FBrsquos manipulations invariably re-spected the physical affordances they did not always leadto correct and efficient use of the objects

In summary of this section we conclude that there is lit-tle compelling evidence to support the hypothesis of an in-dependent component of the semantic system represent-ing action knowledge There is no doubt that the abilityto use objects can be disrupted when conceptual knowl-edge about them is preserved (Rumiati et al 2001 Spattet al 2002) All such reported cases can we think be ex-plained by frank nonsemantic apraxic disorders The twopossible exceptions are the patients studied by Ochipaet al (1989) and Moreaud et al (1998) but in these caseswe question the conclusion that the patientrsquos conceptualknowledge of objects was preserved The other side of theputative dissociation good object use in the face of de-graded object knowledge is a more serious issue We havesuggested above some queries regarding the evidence forthis conclusion in the very few cases in which it has beensuggested but we acknowledge that it remains an un-resolved issue and that the very commonly observed as-sociation (impaired object use consequent on semanticdegraded conceptual knowledge) does not preclude thepossibility of a genuine dissociation Indeed despite ourpreference for a theoretical position that predicts that thisside of the dissociation will not be observed our contin-uing research on the topic is partly motivated by this un-resolved question

AffordancesA 90-object feature database was constructed in order

to identify the systematic relationships between the phys-ical features of an object and the way it is used to assistwith a priori quantification of affordances Affordancewas determined statistically in terms of a consistent re-lationship across items between a structural feature (ega handle of a certain type) and a specific component ofuse (eg a particular type of grip) Despite the size of thisdatabase and the number of possible correlations therewere very few that reached statistical significance Manyof the reliable correlations were either between two differ-ent structural features of an object (eg if the object hastwo handles it is likely to have moving parts) or betweena structural feature and the objectrsquos function (eg if theobject has a sharp serrated edge it is likely to be usedfor cutting) The correlations most relevant to this studyhowever were between a structural feature and the way anobject is held (eg if the object has a handle that joins theshaft it is likely to be held in a ldquostandardrdquo grasp) and be-tween a structural feature and the way an object is moved(eg if the object has a wedge-shaped head it is likelyto be associated with a striking-down movement)

As a group the patients did not achieve better perfor-mance on a subset of affordanced objects when use ofthese was compared with a familiarity-matched subsetof objects lacking such affordances This absence of a

general group benefit applied both to overall use and tothe specific component of use afforded by the objectrsquosstructure When the results were viewed as case-seriesdata with cases characterized by varying degrees of se-mantic impairment however it became clear that therewas a reliable benefit of affordance on the specific com-ponents of use but only for the most impaired patientsThe modulation of affordance by degree of semantic im-pairment follows from the assumptions (1) that object useis governed principally by conceptual knowledge and(2) that affordances have a weak influence on object useThe analyses of the feature database revealed few strongcorrelational affordances whose effects could be detectedonly for the specific component of use It is thereforeonly when semantic memory is severely degraded that onecan readily detect the influence of affordances This pro-posal also explains why we found a familiarity by affor-dance interaction for the most impaired patients The in-fluence of affordances is most obvious for those objectsthat are relatively unfamiliar to the user

Presence of a RecipientIt was hypothesized that having a natural recipient pres-

ent might benefit the patientsrsquo object use in two ways firstby providing a level of context and therefore access to fur-ther conceptual knowledge and second by giving clues asto the ultimate goal (ie the function of the object) andtherefore encouraging trial-and-error problem-solving be-havior The patientsrsquo scores were significantly higher onthe hold of the object and marginally higher on the move-ment when the recipient was present however there wasno effect of recipient on orientation or overall use

The impact of recipient like affordance was found tobe modulated by the degree of semantic impairment Thepatients with a moderate level of conceptual impairmentdemonstrated significantly better use with the recipientpresent whereas the patients with mild and severe impair-ment showed no effect Given that there was little evi-dence for active problem solving in any of these patients(see below) whether or not the object was presented withits recipient it seems most likely that the recipient had itseffect semantically The combination of semantic infor-mation for the object and its recipient could boost perfor-mance but only within a certain range of semantic dete-rioration Two of the mildly impaired patients AN andAT performed close to the normal range on assessmentof single-object use so there was little chance of measur-ing a positive effect when the recipient was present Alsotheir conceptual knowledge was only mildly affected atthis stage so there was little room for improvement In themoderately impaired group the patientsrsquo semantic mem-ory was impaired but the combination of two mildly im-poverished semantic representations (for the object andits recipient) may still be sufficient to constrain objectuse In the most impaired cases however we suggest thatconceptual representations for the object and its recipientwere so impoverished as to prevent any benefit


FamiliarityIt has been repeatedly demonstrated that familiarity is

an important predictor of performance on tests involvingassessment of conceptual knowledge (Bozeat et al 2000Funnell 1995 Lambon Ralph amp Howard 2000) It wasnot surprising therefore to find that familiarity also in-fluenced performance on object use assessments Mostof the patients involved in this study had been sufferingfrom dementia for several years with resulting reductionin the normal variety of daily activities This observationis confirmed by the significant difference between the rat-ings of familiarity obtained from the control subjects andthe caregiversrsquo ratings of how often each patient used the36 objects selected for this study Only personally relevantfamiliarity ratings predicted object use accuracy in thesepatients

There are at least two possible mechanisms by whichuse of familiar objects is maintained First repeated ex-perience with the object may boost degraded conceptualrepresentations which then give the patient enough infor-mation about the object to know how to use it Alternativelythe repeated use of an object may establish a set of auto-matic stereotyped responses that are triggered by thatparticular object and have limited reliance on semanticknowledge These two explanations are not in fact mu-tually exclusive and both may have a role to play

Mechanical Problem SolvingAll the patients performed within the normal range of

control subjects on the Novel Tool test and the mechan-ical puzzles indicating that even the patients with severeconceptual deficits had preserved mechanical-problemndashsolving ability Although it is possible that the Novel Tooltest (Goldenberg amp Hagmann 1998) does not necessar-ily engage mechanical-problemndashsolving skills relyinginstead on visual matching this is not true for the me-chanical puzzles (based on those designed by Ochipa et al1992) Despite this outcome only 2 patients consistentlyused trial-and-error problem solving in the assessmentsof real object use which led to improvements in movementand overall use for one patient (KH) and to no enhance-ment in the other patient (JH)

Why do we see such few examples of problem solvingin real object use even when the recipient is present Wesuspect that the most likely explanation for this again re-lates to the patientsrsquo semantic impairment Without suffi-cient item-specific knowledge the patients are unable toderive the correct function for the object (as corroboratedby impairments on the matching-to-function test) Knowl-edge of function provides the correct goal for the objectwhich is critical for effective problem solving to take placeIt is also possible that knowledge of object properties is re-quired for this level of object use through problem solving(Hodges et al 2000) For example to know that you canturn a screw by using a coin in place of the usual tool youhave to know that the metal will not bend under the twistingforce required One would certainly not try the same thingwith the chocolate coins sometimes given at Christmas

As well as enabling the delineation of the different pro-cesses involved in our everyday interaction with objectsstudies of object use in SD are also relevant to debates onthe streams of visual processing From investigation of theeffects of circumscribed lesions in the macaque monkeyUngerleider and Mishkin (1982) proposed two distinctstreams of visual processing the ventral stream project-ing from the primary visual cortex to the inferotemporalcortex which enables the identification of objects andthe dorsal stream which projects from the primary visualcortex to the posterior parietal cortex and is responsiblefor the localization of objects in space Goodale and Mil-ner (1992) reinterpreted the differences between the twostreams of processing by focusing on the different require-ments of the output systems that each stream serves ratherthan on the different types of information handled Fur-thermore they proposed that skilled appropriate objectuse is possible only through the intact functioning of boththe dorsal and the ventral pathways (Milner amp Goodale1995) Support for the existence of these two streams ofprocessing comes from neuropsychological dissociationsbetween performances on tasks involving identificationof objects and on those involving acting upon them Patientswith optic ataxia who have damage to the superior portionof the posterior parietal cortex are impaired at using vi-sual information to reach out and grasp objects but haveno difficulty recognizing or describing single objects Thepatients described in this study show the opposite disso-ciation They are impaired at identifying objects becauseof extensive temporal lobe pathology but can easily locateand grasp objects in space and are still able to performmechanical-problemndashsolving tasks thanks to the intactdorsal pathway The results of this study therefore sup-port the view that skilled appropriate object use is possibleonly through the intact and probably interactive function-ing of both the dorsal and the ventral pathways

Conclusions The patients with SD involved in this study were im-

paired both on tests of conceptual knowledge and ondemonstrating the use of real objects Furthermore theirdegree of success in object use was significantly corre-lated with their level of semantic impairment providingfurther support for the primary importance of concep-tual knowledge in object use Several other factors havealso been shown to be importantmdashnamely the affor-dances of objects the presence of a recipient and objectfamiliaritymdashalthough in each case this additional influ-ence is modulated by the principal factor the degree ofsemantic impairment

REFERENCES

Bozeat S Lambon Ralph M A Garrard P Patterson K ampHodges J R (2000) Non-verbal semantic impairment in semanticdementia Neuropsychologia 38 1207-1215

Bozeat S Lambon Ralph M A Patterson K amp Hodges J R(2002) The influence of personal familiarity and contexts on objectuse in semantic dementia Neurocase 8 127-134


Buxbaum L J Schwartz M F amp Carew T G (1997) The role ofsemantic memory in object use Cognitive Neuropsychology 14219-254

Buxbaum L J Veramonti T amp Schwartz M F (2000) Functionand manipulation tool knowledge in apraxia Knowing ldquowhat forrdquo butnot ldquohowrdquo Neurocase 6 83-97

Folstein M F Folstein S E amp McHugh P R (1975) ldquoMini-mental staterdquo A practical method for grading the mental state of pa-tients for clinicians Journal of Psychiatric Research 12 189-198

Funnell E (1995) From objects to properties Evidence for spread-ing semantic activation in a case of semantic dementia Memory 3497-519

Funnell E (2001) Evidence for scripts in semantic dementia Impli-cations for theories of semantic memory Cognitive Neuropsychol-ogy 18 323-341

Gibson J J (1977) The theory of affordances In R Shaw J Brans-ford amp N Y Hillsdale (Eds) Perceiving acting and knowing To-wards an ecological psychology (pp 67-82) Hillsdale NJ Erlbaum

Goldenberg G (1996) Defective imitation of gestures in patientswith left and right hemisphere damage Journal of Neurology Neu-rosurgery amp Psychiatry 61 176-180

Goldenberg G amp Hagmann S (1998) Tool use and mechanicalproblem solving in patients with apraxia Neuropsychologia 36 581-589

Goodale M A amp Milner A D (1992) Separate visual pathwaysfor perception and action Trends in Neurosciences 15 20-25

Graham K S Lambon Ralph M A amp Hodges J R (1997) De-termining the impact of autobiographical experience on ldquomeaningrdquoNew insights from investigating sports related vocabulary and knowl-edge in two cases with semantic dementia Cognitive Neuropsychol-ogy 14 801-837

Hamanaka T Matsui A Yoshida S Nakanishi M Fujita KBanno T Murai T Takizawa T amp Hadano K (1996) Cere-bral laterality and category-specificity in cases of semantic memoryimpairment with PET-findings associated with identification amne-sia for familiar faces Brain amp Cognition 30 368-372

Hodges J R Bozeat S Lambon Ralph M A Patterson K ampSpatt J (2000) The role of conceptual knowledge in object use Ev-idence from semantic dementia Brain 123 1913-1925

Hodges J R Graham N amp Patterson K (1995) Charting the pro-gression in semantic dementia Implications for the organisation ofsemantic memory Memory 3 463-495

Hodges J R Patterson K Oxbury S amp Funnell E (1992) Se-mantic dementia Progressive fluent aphasia with temporal lobe at-rophy Brain 115 1783-1806

Hodges J R Spatt J amp Patterson K (1999) What and how Ev-idence for the dissociation of object knowledge and mechanical prob-lem solving skills in the human brain Proceedings of the NationalAcademy of Sciences 96 775-784

Howard D amp Patterson K (1992) Pyramids and palm trees A testof semantic access from pictures and words Bury St Edmunds UKThames Valley Test Company

Humphreys G W amp Forde E M E (2000) Hierarchies similarityand interactivity in object recognition ldquoCategory-specif icrdquo neu-ropsychological deficits Behavioural amp Brain Sciences 24 453-476

Koffka K (1935) Principles of Gestalt psychology New York Har-court Brace amp World

Lambon Ralph M A Graham K S Ellis A amp Hodges J R(1998) Naming in semantic dementia What matters Neuropsy-chologia 36 775-784

Lambon Ralph M A amp Howard D (2000) Gogi aphasia or se-mantic dementia Simulating and assessing poor verbal comprehen-sion in a case of progressive fluent aphasia Cognitive Neuropsy-chology 17 437-465

Lauro-Grotto R Piccini C amp Shallice T (1997) Modality-specific operations in semantic dementia Cortex 33 593-622

Milner A D amp Goodale M A (1995) The visual brain in actionOxford Oxford University Press

Moreaud O Charnallet A amp Pellat J (1998) Identificationwithout manipulation A study of the relations between object useand semantic memory Neuropsychologia 36 1295-1301

Mummery C J Patterson K Price C J Ashburner J Frack-owick R S amp Hodges J R (2000) A voxel based morphometrystudy of semantic dementia The relation of temporal lobe atrophy tocognitive deficit Annals of Neurology 47 36-45

Mummery C J Patterson K Wise R J S Price C J amp HodgesJ R (1999) Disrupted temporal lobe connections in semantic de-mentia Brain 122 61-73

Neisser U (1994) Multiple systems A new approach to cognitive the-ory European Journal of Cognitive Psychology 6 225-241

Ochipa C Rothi L J G amp Heilman K M (1989) Ideationalapraxia A deficit in tool selection and use Annals of Neurology 25190-193

Ochipa C Rothi L J G amp Heilman K M (1992) Conceptualapraxia in Alzheimerrsquos disease Brain 115 1061-1071

Raven J C (1962) Coloured progressive matrices Sets A AB B Lon-don Lewis

Raven J C (1965) Advanced progressive matrices Sets I and II Lon-don Lewis

Rey A (1941) Lrsquoexamen psychologique dans les cas drsquoencephalopathietraumatique Archives de Psychologie 28 286-340

Riddoch M J amp Humphreys G W (1987) A case of integrative vi-sual agnosia Brain 110 1431-1462

Riddoch M J Humphreys G W Coltheart M amp Funnell E(1988) Semantic systems or system Neuropsychological evidencereexamined Cognitive Neuropsychology 5 3-25

Rumiati R I Zanini S Vorano L amp Shallice T (2001) A formof ideational apraxia as a selective deficit of contention schedulingCognitive Neuropsychology 18 617-642

Sirigu A Duhamel J amp Poncet M (1991) The role of sensori-motor experience in object recognition A case of multimodal ag-nosia Brain 114 2555-2573

Snowden J S Goulding P J amp Neary D (1989) Semantic de-mentia A form of circumscribed cerebral atrophy Behavioural Neu-rology 2 167-182

Snowden J S Griffiths H amp Neary D (1994) Semantic demen-tia Autobiographical contribution to preservation of meaning Cog-nitive Neuropsychology 11 265-288

Snowden J S Neary D amp Mann D M A (1996) Fronto-temporallobar degeneration Fronto-temporal dementia progressive aphasiasemantic dementia New York Churchill Livingstone

Spatt J Bak T Bozeat S Patterson K amp Hodges J R (2002)Apraxia mechanical problem solving and semantic knowledge Con-tributions to object usage in corticobasal degeneration Journal ofNeurology 249 601-608

Ungerleider L G amp Mishkin M (1982) Two cortical visual systemsIn D J Ingle M A Goodale amp R J W Mansfield (Eds) Analysis ofvisual behavior (pp 549-586) Cambridge MA MIT Press

Warrington E K (1975) Selective impairment of semantic memoryQuarterly Journal of Experimental Psychology 27 635-657

Warrington E K amp James M (1986) Visual object recognition inpatients with right hemisphere lesions Axes or features Perception15 355-366

Wechsler D A (1981) Wechsler Adult Intelligence ScalendashRevisedTest manual New York Psychological Corporation

Zangwill O L (1960) Lrsquoapraxie ideacuteatorie Nerve Neurology 106595-603

(Manuscript received October 15 2001revision accepted for publication April 12 2002)


relative preservation of general conceptual knowledgesuch as the superordinate category) familiarity personallyrelevant schemata (personal familiarity) and contextsobject affordances and mechanical problem-solvingstrategies As well as having the goal of confirming thatuse is directly tied to the degree of semantic impairmentfor the target object the present study was specificallydesigned to address the role of some of these additionalfactors

Object AffordanceIn 1977 Gibson proposed the theory of affordances in

which he stated that information available more or less di-rectly from the perception of an object gives clues as to itsfunction and possible manipulations (Gibson 1977) Thesearch for the origins of the concept of affordance howevertakes us back to the 1930s and the work of Gestalt psychol-ogist Koffka (1935) who proposed that the meaning orvalue of an object was perceived as readily as its physicalfeatures The term valence was later coined to refer to thisidea There was one crucial difference between the conceptsof affordance and valence Whereas the valence of an ob-ject was thought to be bestowed upon it by the need of theobserver an affordance does not change with the need ofthe observer it is always there to be perceived FollowingGibson the notion of affordance was refined by Neisser(1994) who proposed that the term should be limited tothe notion of physical affordances and not be in any waydependent on stored knowledge about objects In recentstudies the term affordance has been used to refer to twopotentially separate mechanisms that support object usedirectly from visual andor tactile input One is problemsolving or reasoning about the use of an object on the basisof its physical characteristics which will be discussedbelow The other is more like the Gibsonian notion of affor-dance in which clues to the hold orientation movementand purpose of an object are systematically related to itsphysical structure (eg if it has a sharp edge it is used forcutting)

In neuropsychological contexts affordance has typicallybeen characterized in a post hoc fashion If a patientrsquos useof an object is better than would be expected it is assumedthat the physical properties are guiding the patient towardthe correct manipulation For example Hodges et al(2000) reported that the SD patients did use two objects(a pair of scissors and a watering can) better than would beexpected on the basis of their scant remaining conceptualknowledge for these items and suggested that the physi-cal properties of the objects were influencing their useThe construction of a pair of scissors for example is suchthat it is almost impossible not to demonstrate the correctcutting movement In the present study a feature databasewas created not only to improve scoring of object use butalso to identify the systematic relationships between phys-ical features and the way an object is used In this way wehoped to be able to specify affordances a priori before as-sessing their influence on the performance of the SD cases

Mechanical Problem SolvingPatients with SD retain good problem-solving skills

as demonstrated for example by excellent performanceon the Novel Tools test (Goldenberg amp Hagmann 1998)until late in the course of the disease (Hodges et al 2000Hodges Spatt amp Patterson 1999) It is perhaps surpris-ing then that Hodges et al (2000) observed little evi-dence of trial-and-error explorations by the SD patientsthat might have helped them to work out the correct useof real objects The authors proposed two additional con-siderations that may explain this finding First in orderto solve a problem effectively it is necessary to know thetarget or goal The goal is provided by the examiner intests of mechanical problem solving whereas in thedemonstration of single-object use the patients must con-ceive the goal for themselves The ability to deduce thecorrect goal for an object may be dependent on conceptualknowledge (ie the patient must know what the objectrsquoscanonical function is) Second the ability to work out aplausible function for an object often requires knowledgeabout its properties (eg knowledge about the materialsit is made from) which may also be impaired in thesepatients A further possible factor is that the mechanical-problemndashsolving test used in the previous study the NovelTool test (Goldenberg amp Hagmann 1998) can be accom-plished purely through visual matching of the tools andthe blocks making it a nonstringent measure of real prob-lem solving

Several aspects of the present study were designed inorder to explore these possible explanations for the appar-ent absence of problem-solving behavior in real objectuse The patients were assessed on single-object use withand without a recipient (see below) to investigate whetherthe presence of the recipient provides clues as to the goal(the objectrsquos function) thereby enabling better problemsolving We also included a more difficult assessment ofmechanical problem solving mechanical puzzles (basedon those designed by Ochipa Rothi amp Heilman 1992)which cannot be accomplished simply on the basis of vi-sual matching

Presence of a RecipientPrevious studies by our group (Hodges et al 2000

Hodges et al 1999) have assessed use of single objects inisolation to enable investigation of the role of conceptualknowledge without contaminationby other factors that areinvolved in object use Obviously this does not accuratelyreflect our everyday interaction with objects where wewould typically use two or more objects together to carryout a task Indeed this may go some way toward explain-ing how the SD patient reported by Lauro-Grotto Picciniand Shallice (1997) was able to prepare and cook a varietyof different foods despite performing poorly on tests assess-ing conceptual knowledge of the same items We proposethat the recipient may benefit the patients in two waysFirst it may provide a level of context and therefore helpto constrain the conceptual knowledge that underpins cor-







METHOD













Control


Age 62 59 65 58 61 59 78 70 637 62


























RESULTS













Subject Control




Living 32 25 27 23 15 10 9 3 318 04Manmade 32 31 30 28 25 8 14 5 320 02




Subject Control


MMSE (30) 30 24 25 22 12 7 8 7 288 05Digit span



























































DISCUSSION








































REFERENCES























































METHOD













Control


Age 62 59 65 58 61 59 78 70 637 62


























RESULTS













Subject Control




Living 32 25 27 23 15 10 9 3 318 04Manmade 32 31 30 28 25 8 14 5 320 02




Subject Control


MMSE (30) 30 24 25 22 12 7 8 7 288 05Digit span



























































DISCUSSION








































REFERENCES






































































RESULTS













Subject Control




Living 32 25 27 23 15 10 9 3 318 04Manmade 32 31 30 28 25 8 14 5 320 02




Subject Control


MMSE (30) 30 24 25 22 12 7 8 7 288 05Digit span



























































DISCUSSION








































REFERENCES



























































Subject Control




Living 32 25 27 23 15 10 9 3 318 04Manmade 32 31 30 28 25 8 14 5 320 02




Subject Control


MMSE (30) 30 24 25 22 12 7 8 7 288 05Digit span



























































DISCUSSION








































REFERENCES







































































































DISCUSSION








































REFERENCES















































































REFERENCES

















































When objects lose their meaning: What happens to their use?

Documents

Transcript of When objects lose their meaning: What happens to their use?