Neural mechanisms of timing

Hazeltine

et

al.

Neural mechanisms timing Eliot

Hazeltine,

Laura

L. Helmuth

pmagmss ii&n@ &borate temp@@

experimental

sensory

and mnemonic

characteristics

of the clock

@m&&gsystem. s,@th

Similar

B/s. However,

other

patient

P

recise timing

t

from

clinical

are impaired

at discriminating

populations,

especially

deficits

of activity across the muscles of the shoulder, arm and wrist. events, such as when a traffic light will

stud and

those

wit#) d

in ti throulgti:~:Q~! :j i >,-l_ _,(;,:j::,_I “:L,iLp,I)*‘i”_q.,, /<.)Ljq ./ $.)ii$i@ i ),j i x;i ‘y :(i I ‘j$‘$:i::::_ : $.!& (, ‘.‘. (:::; :,:fil:)::‘i ;irlrlii)‘-:‘Ir~:-‘~.,i’+;: :fiq

may be distributgd

is essential in many of our behaviors.

predicting

b&

can be reached

Reaching for an object requires a specific temporal pattern Similarly,

B. Ivry

an

in perception

damage

mechanisms

Mark

processes.

#a, also exhibit v

designs

conclusions

neocerebellar

made

Timing

of

and Richard

has been

-

when we enter a nightclub,

we immediately

find ourselves

moving to the beat. Along this line, Treisman and colleagues examined how trains of evenly spaced clicks affected time perception’ production

and

task.?. In both cases, the influence of the clicks

change to red, demands an accurate estimate of relevant durations, such as how long the light has been yellow.

was shown to be dependent on their frequency;

These examples emphasize that temporal information underlies both sensory and motoric processes; in everyday

frequencies systematically increased or decreased perceived or produced intervals. Moreover, frequencies that shortened

activities,

estimates of perceived time, presumably

temporal

intervals

must be perceived

and pro-

by slowing down

the clock, tended to lengthen movement times, and vice versa. This significant negative correlation between the biases

duced correctly. Component

particular

analysis of temporal

For both perception and production

variability

observed

there is, of course, vari-

ability in temporal accuracy. The components of a movement differ from trial to trial, and on perception tasks, we make slight misestimations of elapsed time intervals. However, errors can also emerge from a variety of sources that do not necessarily involve timing per se but, instead, result from variability in other task-related processes. Indeed, movement commands may be initiated at the correct time, but delays in implementation may occur downstream, leading to poor performance. An analogous situation exists for

on the perception

and production

tasks is consist-

ent with the hypothesis of a common timing mechanism. Ivry and Hazeltine performance

also found similarities

of time perception

between the

and time production

intervals by tapping and made temporal judgments across a range ofdurations (325-550 ms). The increase in the variance as a function of the duration was comparable for the two tasks. Assuming that this duration-dependent component of the variance provides an estimate of variability in an internal timing mechanism, these results implicate a common timing

perception. Returning to the example of the traffic light, an error in predicting the change to red can be caused by mis-

process in the two tasks. Studies measuring variation

judging

hypothesis. Keele eta/.* observed significant

the relevant duration

or by failing

to detect the

using

the slope method (see Box 1). Subjects produced temporal

individuals’

perceptuomotor

among

skills have also supported this correlations

be-

light’s transition from green to yellow rapidly enough. Box 1

tween performance in a repetitive tapping task and a duration

describes two approaches used to partition

discrimination

temporal tasks into component

variability

in

sources.

acuity on the perception

motor task (see also Ref. 5), suggesting that the correlation reflects an overlap in the neural systems subserving these tasks.

A common mechanism Are perception

task. Furthermore,

task was not correlated with performance on a non-temporal

and production

subserved by the same

clock? Evidence for a common timing mechanism can be observed in the tendency for movements to become en-

Neural substrates

trained automatically

to look for converging

with external stimuli.

Copyright

For instance,

0 1997, Elsevier

Science

Trends

Given the correlational

Ltd. All rights in

Cognitive

reserved.

nature of these studies, it is important evidence supporting

1364.6613/97/$17.00

Sciences

-

Vol.

the common

PII: 513646613(97)01058-9 1,

No.

5,

August

1997

Hazeltine

et

al.

-

Timing

mechanisms

Box 1. Isolating Wing and Krisrofferson’ variability tinued

developed

on a motor

their movements

timing

timing

variability

nated. Variability

movements

independent.

Secondly, rhe operation

OCCUI in an ‘open-loop’

duration-dependent

correct errors on previous implementation

variance reflected the operation ofan internal or motor im-

should remain constam across different durations.

Close agreement is found for estimates of implementation

of these processes must is made to

ability

mean that

Kristofferson

taps. These assumptions

variability

plementation,

systems must be

fashion so chat no attempt

processes, defined by the

clock, as other processes, such as signal detection

for this model. Firstly, the vari-

ability of the clock and motor implemenration

and durarion-independent

slope and inrercepr rcrms, respectively. It was assumed rhar the

imple-

system that rranslared this signal into a movcmenr.

Two assumprions arc important

motor

ccpt. Thus, regression analyses were used to separate durationdependent

an appro-

and a motor

this logic in

target durarion was very nearly linear and had a positive y-inrer-

during the unpaced phase was hypothesized

rimed trigger for each movement

Ivry and Hazeltine’applied

task in which subjects capped over a range of

time intervals (Fig. b). The function relating the variance to the

after the signals termi-

to arise from two sources: an inrernal clock providing mentation

a time production

subjects matched

to a pacing signal (e.g. 2.5 Hz), and then con-

to make periodic

p&rely

heres to Weber’s la+.

a simple model to analyze

task. Initially,

from the slope method

calculated

vari-

or via the Wing

and

model (see also Ref. g).

causes a negative correlation

between the timing of successive inrervals whereas the effects of clock variability

are restricted

Thus, implemenrarion

co a single interval

variability

a Wing,

the co-

timing

can be identified

variance between successive intervals. mate from rhe total variability,

(see Fig. a). by

By subrracting

in the time perception

duration

being estimated.

literature:

density:

implications

Behav.

Proc. 23,44-55

d Getty,

of the square of rhe

In other words, time perception

A. (1973) responses

The perception

c Bizo, L.A. and White,

The slope method is an alcernarive approach thar is based on a phenomenon

L. (1979)

motor

Response

Percept

delays

Psychophys.

of time

Percept.

and

the

14, 512

Psychophys.

26.

340-354

an estimate can be made of clock

rhar variance increases linearly as a function

Kristoffenon.

of discrete

b Allan.

this esti-

variability. well-established

A. and

D. (1975)

comparison

ad-

e Gibbon,

K.G. (1997) Timing for models

Discrimination

of two

models

1. (1991) Origins

with

of timing

of short

Percept

controlled

temporal

Psychophys.

of scalar timing

reinforce

J. Exp. Psycho/.

Anim.

intervals:

a

18, l-8

Learn.

MO&.

22,3-38

Time c Central process

C

C

C

I \

I \

C

I \

C I \

I \

C

C+D I \

C

I \

I \

I

Peripheral process

Interval

I

Overt response Fig.

Wing

about

independently;

Kristofferson

half

defined

by that Note

that

in one have

of figure), tap,

the

will

in a typical

no effect

interval

whose

be shortened. tapping

Tapping

process

on the progression

clock hypothesis. Neuropsychological

the

interval

Tap

is attributed

to two

influence

C is much

processes. each

be lengthened

of clock

variability

is limited

the

motor

delay

The central

is delayed

tap will

than

Tap

command.

If implementation

by the

longer

I

I+D

Tap

M ms to implement

of the other.

is defined the

Tap

on average

termination

In contrast, task,

variability

requires

I

I

Tap

model.

C ms. The peripheral

delays

(see left ure).

and

every

I-D

Tap

Tap

a The

mands

I+D

and the to a single

Tap

process

The two

by D ms in the next

interval,

interval

issues tap

processes motor whose

(see right

com-

proceed

half

process onset

is

of fig-

M.

research provides just

pathology ” . Patients with medial cerebellar lesions suffered

such an alternative method for investigating the neural bases of internal timing. The cerebellum has been shown to

from marked peripheral (motor) variability, while those with lateral cerebellar lesions showed impaired central

be critical for a wide variety of timing tasks. Some of the cardinal symptoms of cerebellar dysfunction (dysmetria and

(clock) variability.

intentional

tremor) have been attributed

to a loss in coordi-

Importantly,

temporal

deficits

following

cerebellar

mus-

damage have been reported in the perceptual domain as well. Ivry and Keele” observed that such lesions not only in-

cle&‘. Cerebellar dysarthria is most evident on sounds that require precise timing between different sets of ardculator?.

creased clock variability during tapping, but also impaired performance on a duration discrimination cask. The pa-

Patients with cerebellar lesions also show increased variabil-

tients’ performance was comparable to control subjects on a

ity on a repetitive

loudness

nation of the temporal

pattern between antagonist

tapping rask9, and when these data are

analyzed with the Wing and Kristofferson a dissociation

Trends

in

is observed as a function

Cognitive

Sciences

- Vol.

model (see Box I), of the locus of the

1,

No.

5,

August

discrimination

task,

demonstrating

deficits were not related to the perception to general factors such as motivation.

1997

that

their

of the stimuli or

Cerebellar

patients

f Ivry, R. and Hazeltine, temporal

intervals

common

timing

Perform.

RX. (1995) across

a range

mechanism

Perception

and production

of durations:

J. Exp.

evidence

Psycho/.

Hum.

of for

a

Percept.

A.

(1980)

sequences, Requin,

in

tion task. While Ivry and Keele9 failed to find performance differences

21,3-18

p Wing,

Structures within the basal ganglia have also been implicated in timing functions. Jeuptner etaLL3 reported rCBF increases in the basal ganglia (as well as cerebellar temporal, prefrontal and cingulate cortex) during their time percep-

The

long

and

Tutorials

in

Motor

1.. eds). pp. 463-486,

short

of

timing

Behavior

in

(Stelmach,

between a group of Parkinson’s

controls on the repetitive

response G. and

North-Holland

have found

that patients

with

either

Huntington’s

or

Parkinson’s

disease have more variable inter-tap intervals than control subjects r4,15.Moreover, the performance of the patients with Parkinson’s disease was improved by L-dopa

A

800

medication,

and those with asymmetric symptoms produce

more variable time intervals with the more affected limb. Decomposition of the time interval variability using the

s 600 5 8 400 s ‘C

Wing and Kristofferson”j increased temporal and peripheral Cortical

9 200 0

I

I

325

400

the

to both the central (clock)

(motor) sourcesIs. structures

those attributed

1

model (see Box 1) attributed

variability

haps more integrative

seem to contribute computations

different,

in timing

per-

tasks than

to the cerebellum or the basal ganglia. For

I

475

550

example, premotor

or supplementary

motor cortical lesions

have been associated with deficits in rhythm production”. Furthermore, laterality effects have been reported on tests of

Duration (m$) B

patients and

tapping task, more recent studies

temporal reproduction:

800

for intervals ranging from 1 to 5 s,

patients with precentral left hemisphere lesions tend to produce shorter time intervals than control subjects, whereas patients with precentral right hemisphere lesions tend to

$- 600 .c.

produce longer time intervals’s.

Component analysis of the neural systemsfor timing: animal models Reports of timing deficits in many different

patient groups

have led some researchers to conclude that temporal

325

400

475

550

production the

(0) and perception

variance

Regression

was lines

plotted show

was presented

once

(6) The target

interval

both

experiments,

the

and perception

tasks.

affected

number

by the

suggests there

that

best

in both

for both

fit.

slope Moreover,

of intervals

presentations

duration

when

four

and

perception

times

in both

magnitude produced variability

of the target

for the

interval tasks. tasks.

In

production

of the slope or presented. is reduced

Indeed, theoretical

models of neural clocks typically

discrete components each ofwhich

have

is envisioned to perform

a specific operation2,‘9.

squared.

(A) The target

was comparable the

the time

best described

of the

the production

duration-dependent

are multiple

linear

was presented

the

for time percep-

tasks (0) were

as a function

pro-

across a variety of cortical and subcor-

tical systems I5. However, the plurality of participating brain regions does not necessarily imply functional homogeneity.

Duration (ms2) Fig. b Evidence for a common mechanism tion and production. The variability functions

cessing is distributed

Pharmacological helped to differentiate

and lesion studies with rats have the functions of the cortical and sub-

cortical areas implicated in temporal processing. One widely-

was

used task measures the frequency

This

(bar presses) over time. During training trials, reward is pro-

when

interval.

have also been found to perform poorly on tests of motion and velocity perception”,‘*, tasks that may rely on a representation of precise temporal information. Furthermore, bilateral increases in regional cerebral blood flow (rCBF) were observed with positron emission tomography (PET) when

of an animal’s response

vided for the first response that is made after a set interval has expired (e.g. 30s). In this paradigm the rats learn to withhold responding until after the set interval has expired, despite the fact that the end of an interval is not cued explicitly.

The response rate typically

increases when the rats

judge that the interval has expired. Performance is then evaluated on trials in which no reward is given. Drugs that target dopaminergic

and cholinergic

systems produce a shift

subjects judged the duration of tone inrervals’3. While these

in the time at which the response rate peaks. Importantly,

imaging results are in accord with the cerebellar timing hypothesis, the activation can be attributed to a variety of

these drugs appear to change temporal processing via differ-

sources. Indeed, the timing condition

dramatic but diminish

was compared with a

ent mechanisms. The dopamine-related with prolonged

effects are initially drug exposure. This

control task in which subjects were not required to make any sensory judgments. Thus attention, motor selection

pattern is consistent with the idea that the drug changes the speed of an internal clock and over time the animal is

and other cognitive

able to modify

demands may have contributed

to the

activations observed.

its memory of the reinforcement

contrast, the acetylcholine-related

Trends

in

Cognitive

Sciences

-

Vol.

time. In

effects develop slowly but

1,

No.

5,

August

1997

30 s (Ref. 23). Human studies also suggest a qualitative change in temporal processing for intervals longer than 1.5-2 s (Ref. 24). When attempting to tap in synchrony with a series of tones, responses are anticipatory when the inter-tone interval is less than 1800 ms; for longer intervals, the responses tend to be reactive to the tones25. One might hypothesize that the cerebellum is critical for short intervals and other structures, such as the basal ganglia and cortex, take over for longer intervals. However, it remains possible that as the target interval lengthens, additional

pro-

cesses involved in memory and attentional functions come into play. It is important Fig. 1 A three-component striatum

to the

pallidurn.

is sent via the thalamus rations.

This model

It is not

clear

how

clock-counter where

the internal

to the frontal

has been

segment

cortex,

proposed

such a system

system. The substantia where

to account

could

time

the

nigra

sends regularly

(GPi) acts as an accumulator. it is compared for data

activations

from

with animal

of different

stored

paced

Efference

from

representations

studies

using

muscle

groups

signals

through

this accumulator

of task-relevant

durations

the du-

of 20 s or longer”.

in multijoint

movements,

to note that the majority

of the

animal timing studies have not included non-temporal

control tasks. For example,

do dopaminergic

agents distort

perfor-

mance when the task requires discriminations along a dimension such as stimulus intensity!

Studies

along

this line

are more long lasting, suggesting that they affect the ani-

would help develop a component

analysis of temporal pro-

mal’s stored representation of the target interval (reviewed in Ref. 20). Meckz” proposed a multi-component model to account for these findings (Fig. 1). A dopamine-dependent, basal

cessing tasks, and in particular,

help identify

ganglia system forms the pacemaker-accumulator mechanism, and an acetylcholine-dependent, frontal cortex system is associated with Moreover,

within

for further

temporal

memory

and attention.

the basal ganglia system, there is evidence

specialization.

Lesions to the substantia

(SN) prevent rats from learning temporal

nigra

discriminations

when the time interval is more than 20 s but less than 60 s,

if a target

neural system is linked specifically to temporal processing. This strategy has been explored in a recent study with humansZ6. Cerebellar and frontal patients were tested on duration discrimination tasks and pitch discrimination tasks in which the interval between the standard and comparison interval varied between short and long. Cerebellar patients were impaired on both duration discrimination

tasks (inter-

vals centered on 400 ms and 4 s), while frontal patients were impaired on both tasks, but only when the stimuli were separated by 4 s. This result suggests that the frontal contribu-

whereas damage to the dorsal striatum disrupts performance

tion on these tasks may not be specific to timing, but may

predominantly on the 60 s intervals. To explain this dissociation, Meek hypothesized that the SN provides a time-

manifest whenever the working

keeping

pulse which

the dorsal striatum

integrates

for

memory or attentional

re-

quirements are increased. A further problem is that investigations into the role of the cerebellum and basal ganglia in timing have emphasized

longer interval discriminations. Output from this integrator is compared with stored representations of the target interval via the frontal lobes. This model is also consistent with

different dependent variables. Timing abnormalities following cerebellar lesions are manifest as increased variability,

the timing

whereas basal ganglia research has focused on changes in

behavior observed in patients with Parkinson’s

disease, who tend to underestimate date, the cerebellum framework Do distinct

time intervalsa’.

has not been considered

within

To

bias. The former effects have been viewed as evidence of a

the

noisy timing system, the latter as a change in clock speed.

of this model.

A multi-component

mechanisms operate at different

temporal

mechanism

can produce

such deficits in a variety of ways. According

timing

to the model

presented by Meckzo, the dorsal striatum acts as an accumu-

intervals?

lator that stores output from other sources. A related idea

With some exceptions’5, research examining the role of the basal ganglia in timing has involved time intervals spanning

builds upon the hypothesis that the basal ganglia are critical for shifting cognitive set2’ 30.Patients with Parkinson’s dis-

several seconds or more2K22, while studies focusing on the cerebellum have used events of less than 1 s (Refs 7,9). Does this mean that these two systems mediate timing at two dif-

larly when both tasks use the same stimuli. In such tasks, the internal ‘set’ as well as the external stimuli determine the

ease take longer to switch from one task to another, particu-

ferent areas along a temporal spectrum? Some animal research supports this hypothesis. For example, cerebellar lesions in

correct response. If timing long durations, in the absence of

the rat led to a selective deficit on a duration discrimination

and the updating of internal states, then dopamine depletion

task when the stimulus range was centered at 500 ms, but

may slow this process, leading

did not affect performance

parkinsonian

Trends

in

Cognitive

when the range was centered at

Sciences

-

Vol.

1.

No.

5,

August

external cues, requires the accumulation

1997

of short intervals

to timing

problems

patients. The pacemaker providing

in

the short

Hazeltine

Box 2. How a cerebellar

timing

system

Cerebellar Purkinje cells are the site of a tremendous convergence of information; a single Purkinje cell may receive inputs from as many as 200 000 parallel fibers. This architecture is ideal for pattern recognitior@. These cells might learn to recognize the duration-dependent pattern of activity along rhe parallel fibers to signal when expected events should occur. Input to Purkinje cells could code duration in several ways. For example, granule, stellate and basket may interact to operate much like the networks transducing temporal information to spatial information described by Buonomano and colleagues’,d. According to this explanation, relatively slow pre- and postsynaptic mechanisms allow neurons to create local representations of temporal information. For example, Buonomano and Merzenichd have shown how temporal estimates can be derived from a neural network with random variation in the time course of paired pulse facilitation and slow inhibitory posrsynaptic potentials. Because of these properties, the network is in a continuously changing state after firing in response to an initial stimulus. Therefore, responses to subsequent stimulation are contingent on the length of the intervening interval. A second hypothesis proposes that the cerebellum may house populations of interval-based timers, each coding a different duration for a particular effector’. Just as in the visual cortex, in which specific neurons code for orientarion in each region of the visual field, units in the cerebella cortex could code for a specific time inrerval. These units could then be associated with specific muscle groups, or be linked with perceptual infor-

intervals could be instantiated

by other structures such as

might

et

al.

-

Timing

mechanisms

operate

mation. When more than one effector is used,such aswhen an individual taps with both the left and right index fingers, output from all the units is averaged, resulting in a decreasein variabili$. These two accounrs are not mutually exclusive. Such timing mechanisms are unlikely to be adequate for intervals much greater than 1 s. Longer durations may be timed by the striatal system described above, or by the cerebellum in concert with orher neural suuctures. For example, the cerebellum might time srate changes in the frontal cortex with the basal ganglia required to effect these state changes. In this model, the cerebellum could provide the timed initiation signal while the frontal strucmres act as an accumulator. References a Albur.

J.S. (1971) A theory

of cerebellar

function

Math.

Biosci.

10,

25-61 b Marr,

D.A.

(1969)

A theory

of cerebellar

cortex

1. Physiol.

Neural

network

202,

437470 c Buonomano,

D.V. and Mauk,

of the cerebellum d 8uonomano,

D.V.

information with 8 Ivry,

f Helmuth, one:

and

The

(1994)

M.M.

a spatial

representation control

of temporal

Curr. Opin.

L.L. and Ivry. R.B. (1996) When

1. Exp. Psycho/.

temporal

timing Hum.

(1995)

Temporal

code by a neural

network

Science 267, 1028-1030

and motor

reduced

model

6,38-55

Merzenich, into

properties

R. (1996)

M.D.

Computat

transformed

realistic

perception

Neural

variability Percept

information

for

in

6, 851-857

two hands are better

during

Perform.

information

Neurobiol. bimanual

than

movements

22, 278-293

the

sensory

and

motoric

the cerebellum or substantia nigra, but the critical represen-

processes. Strong evidence for the cerebellum’s ability to as-

tation of time would be related to the updating or changing

sociate inputs from different

of behavioral stateP.

experiments establishing its critical role in eyeblink conditioninti5, a learning task in which there is a need to encode

How the cerebellum might perform timing Beginning with Braitenbere33, many theorists have sug-

the precise temporal relationships between the unconditioned stimulus and the conditioned stimulus36,37. Perhaps

gested that the cerebellum’s architecture is well-suited for calculating the precise temporal relationship between differ-

when there is only one source of input. For example, when

ent inputs and between input and output patterns (see Box 2). In a recent review, Raymond etaL3* argued that the

the duration of a short tone must be judged, the cerebellum may compare its representation of the tone with some

this system has evolved so that it can continue

cerebellum plays a critical role in timing based on the com-

internal

putational

signal.

overlap

between two behaviors

for which

the

structure has been shown to be critical, eyeblink conditioning (reviewed (VOR).

in Ref. 35) and the vestibular

ocular reflex

In both cases, motor outputs (e.g. eyeblinks or eye

sensory channels comes from

to operate

standard instead of .a second stimulus

or motor

In motor control theory, it has been proposed that the cerebellum

coordinates

performing

complex pattern recognition

and fine-tunes

cortical outputs by across a wide range

movements) must be timed precisely to be effective.

of inputs 38,39.This sort of computation

Why should the cerebellum, traditionally labeled as a motor structure, be involved in the perception of time? In

pret time-varying representations of sensorimotor patterns so as to estimate elapsed durations4”,41. In other words, the

many cases, the distinction between time production and time perception is unclear. Consider, for example, when one attempts to grab a moving object, a task widely consid-

cerebellum may train other brain regions to recognize and

.ered to be the purview of the cerebellum. Here, motor com-

tervals. This hypothesis has some similarity

mands must be integrated with dynamic sensory information to perform the action properly. Timing is relevant

posed by Courschesne and colleagues42,43, who suggest that the cerebellum coordinates internal operations with antici-

for both predicting

pated sensory information.

the location

of the target object and

can be used to inter-

anticipate representational states by identifying activations across sets of neurons that are associated with particular inwith that pro-

Discrete motor actions are gen-

scheduling the activation of component muscle groups. A system that learns to solve such complex problems

erally less than 1 s in duration, and the hypothesized mecha-

might

be limited

analyze correspondences

events rather than independently

between

the two sets of

compute

the requisite

Trends

nisms for sustaining local neural activity are also thought to to relatively

short durations*‘.

Thus, for longer

durations, such a system is unlikely

to be sufficient.

in

No.

Cognitive

Sciences

-

Vol.

1,

5,

August

1997

7 How.

Outstanding

questions

1.. Wild,

elbow, 8 Ivry.

l

l l l

Evidence suggests that perception and production share a common timing mechanism. How is such a system implemented? How do ensembles of neurons provide accurate temporal information? What are the temporal ranges of these ensembles? How do the basal ganglia, cerebellum and frontal lobes interact to perform temporal computations?

B. and Diener,

wrist,

and fingers

R. and

patients

Gopal.

with

Cognitive

H. (1992)

cerebellar

XIV: Synergies

in Experimental

medial

cerebellum

with

M. and cerebellar

a pacemaker or oscillator44. Moreover,

13 Jeuptner,

Wing and Kristofferson

the bimanual

condition.

When

the advantage was restricted to the clock component of the variability. The researchers interpreted this advantage as evidence that independent clocks subserved the two fingers, and that the clock signals were averaged to determine movement initiation

time. In the present framework,

provement would be attributed

Temporal error takes many forms, which suggests that multiple operations are required to compute time. The results of human and animal studies suggest that different neural systems are essential for performance in timing tasks. The

with

is distributed

multiple systems, a more parsimonious tem makes a distinct contribution.

across

the use of a

motor

U. area

16 van

Steinbuchel.

oscillator

underlying

2 Treisman,

time

a temporal

timing 4 Keele.

controlling

Brain

Hazeltine,

J. Exp. Rychol.

timing

mechanism?

J. Motor M., with

of motor timing

On the

action:

21 Pastor,

M.A.

abnormal

Behav.

Shahani, cerebellar

its

relation

22 Maricq,

A.V.

haloperidol

1.

Perception

Hum.

Percept. and

Acta Psycho/. R. (1987)

19,9&l

B. and

and

production

evidence

in

and

R. (1975)

J. Neural.

EMG

Neurosurg.

analysis

Psychiatry

Sciences

-

Vol.

1,

No.

5,

Brain

E. (1996)

116,

Timing

of the brain, and Artieda,

in

in Time,

J., eds), pp.

of time

J., edr), pp. 143-164,

J.

et

internal

a/.

synchronization

Cognition

perception

The

differential

effects

of

estimation

in the

rat

timing M.A.

and

integration

in

sensorimotor

6. 332-340

at the

Society,

manifestations Fourth

March

of a cerebellar

Annual

Meeting

of the

23.

CD. (1988) Internal in Parkinson’s

J.J. et a/. (1993)

(Pastor,

of Time, Harper

1. (1997) The many

of attention

of the cerebellar

Movement

Temporal

R.G. and Marsden.

is

Elsevier

Presented

Neuroscience

reproduction

on time

and

1. Cogn. Neurosci.

mechanism.

and

115. 21 l-225

the domain

Clocks,

(1994)

26 Ivry, R. and Mangles,

Brain

and time

estimation

(1983)

Exploring

Artieda,

versusexternal

disease Brain

Alternating

fluency

cues and

111, 323-345

in Parkinson’s

disease

116,887-902 A. et a/. Toward

a functional

(1994) Different

and internal

cues in Parkinson’s

P.R. and Fetterman,

psycho/.

analysis

of the

basal

ganglia

(in press)

30 Fimm. 6. eta/.

mechanisms

underlying

disease Brain

Cognit.

J.G. (1988) A behavioral

set on external 25. 287-304 theory

of timing

Rev. 95,274-295

32 Braitenberg,

V. and Atwood,

33 Braitenberg.

cortex

R.P. (1958)

J. Comp.

Neural.

V. (1983) The cerebellum

J.L.. Lireberger,

a neuronal

of 38,

learning

J.J. and

mechanisms

August

R.M.

in Time,

Neurosc;.

Young,

Church,

system,

35 Kim, and its relation

14

lesions

Time

24 Fraisse, P. (1963) The Psychology

a

60, 173-193 Force control

Poeppel,

M.A.

disease Brain

5. et a/. (1996)

34 Raymond,

share

in man

Morphological

observations

109. l-34 revisited

J. Theor.

Neurobiol.

2,

237-241

21,3-18

production

(1992)

79, 10-l 5

temporal

1997

S.G. and Mauk,

machine?

Thompson, involved

M.D. (1996) The cerebellum:

Science 272. 1126-1131

R.F. (1997) in

classical

Cerebellar

circuits

eyeblink

conditioning

and synaptic Trends

20, 177-181

36 IVV. R. (1993) Cerebellar

37 Perret.

Cognitive

a/.

Psychopharmacology

25 Mates.

for a common

Perform.

motor

of

and the supplementary

disturbances

(Pastor,

methamphetamine

31 Killeen,

Quart

and the timing

14. 5-12

of movement

M. and

and

23 Clarke,

for

1163-1169

Trends

et

1. Cogn. Neurosci.

of

control

tasks after

in Parkinson’s

the control

evidence

of movement

of durations:

Do perception

S., Ivry, R. and Pokorny,

patients

P.L.N. (1992)

the

R.E. (1995)

mechanism

5. et al. (1985)

6 Hallett,

estimates

delays

Rychophys.

Res. 3. 217-242

on the cerebellar

across a range

to timing

Naish,

and the timing

intervals

common 5 Keele.

some

Response

Percept.

of timing

Cognit.

45,235-263

R. and

temporal

with

and

in subjects

119, 51-70

20 Meek. W.H. (1996) Neuropharmacology

19, 705743

A. and

perception oscillator

Exp. Psycho/.

perception

Perception

M.. Faulkner,

between

3 Ivry,

time

frequency

in Huntington’s

movements

Representation

29 Hayes,

characteristic

processes

F.W.J. (1996) The accuracy

R.M. (1990)

27 Brown,

for a temporal

timing

timing

repetitive

A. (1973)

N., Wittman,

J. and Church,

28 Downea,

evidence

from

37,23-54

. . ... .. ... .... . .. . .. .. . .. .. ... .... . . .. .. . . .. ... .. .... . .... .. ... ... ... .. .... ... clock:

deficits

Elsevier

19 Gibbon,

of the relations between these neural

M. et a/. (1990) The internal

of motor

Clocks, and Movement

plete understanding

1 Treisman,

perception

(1993) The role of premotor

and motor

Internal

Cognitive

References

in patients

3, 355-366

of cerebellar

J.S. and Cody,

responses

eta/.

perception

Vis. ttes. 35, 723-731

of self-paced,

in the temporal

perceptual

timing

processes.

Exp.

243-266

range of tasks and temporal intervals is essential for a comsystems in both temporal and non-temporal

and

execution

545

disease Brain

motor

17 H&band,

view is that each sys-

Therefore,

1. Cogn.

Dis. (in press)

A. and Kristofferson.

of discrete

velocity

Localization

R&t.

of timing

Parkinson’s

16 Wing,

question that remains is how do we characterize the functions of these different systems? While it is possible that the of temporal information

pp.

of the lateral

and movement

Motion

in human

45, 1540-l

D.J.. Freeman,

precision

to a gain in the information

Conclusion

and

S., eds),

of the cerebellum

1. Cogn. Neurosci.

J.S. et a/. Abnormalities

281-304,

dynamic properties than the signal from just one finger.

timing

M. (1995)

lesions

M. et a/. (1995)

15 D’Boyle,

the im-

available. The signal from the two fingers would have richer

representation

Rizzo,

PET Neurology

14 Freeman,

analysis was applied to these data,

functions

H.C. (1991) Impaired

disease Parkinsonism

and Ivry4’ compared unimanual and bimanual finger tapping, and observed that the temporal variability of each finger was reduced during

using

In addition, Helmuth

Intelligence,

Kornblum,

H.C. (1988) Dissociation

lesions of the cerebellum

midline

form of computation.

in

Volume

Res. 73, 167-180

12 Nawrot,

non-oscillatory

Artificial

D.E. and

in movement

11 Ivry, R. and Diener,

ture. Indeed, despite the hypothesis that the cerebellum is critical for timing, there is a little evidence that it behaves as the character of cer-

production

1, 136-l 52

10 Ivry. R.. Keele. S. and Diener.

ebellar timing deficits (an increase in variability without a bias to shorten or lengthen intervals) is in accord with a

and

MIT Press

Neurosci.

Brain

at the

and Performance

Psychology,

(Meyer,

dysmetria

perception

in Attention

9 Ivry, R. and Keele. S. (1989) Timing

A pattern recognition approach to timing could help account for some notable phenomena in the timing litera-

Cerebellar 65, 563-571

Speech

lesions,

Neuroscience

771-802,

H.C. (1991)

J. Neurophysiol.

information 5.. Ruiz,

involvement Ann.

B. and

New Mauk,

in the explicit

representation

of

York Acad. Sci. 682, 214-230 M. (1993)

Cerebellar

cortex

lesions

Hazeltine

disrupt

learning-dependent

1. Neurosci. 38 Houk,

J.C. and Barto,

Tutorials

in Motor

71-100, 39 Thatch.

timing

of conditioned

eyelid

properties

responses

42 Allen,

13, 1708-1718 A.G. (1992) Distributed

Behavior

/I (Stelmach,

sensorimotor

learning,

G.E. and Requin.

1.. eds). pp.

functions

W.T. (1996)

learning

On the specific

and cognition:

man Behav.

Brain

40 Euonomano,

clues from

Neural

41 Buonomano, transformed

44 Keating.

in motor

and lesion

studies

olive

in

Computat.

M.D. (1994) Neural

model

M.M.

into

by a neural

code

(1995) Temporal network

Timing

mechanisms

G. Prediction

J.G. and Thach.

Learn. W.T.

interspike

in the awake

activation

(1995)

behaving

the

and preparation,

Mem.

interval

of

cerebellum

275, 1940-1943

Science

fundamental

(in press) Nonclock

of

behavior

Purkinje

monkey

cell

is random

of inferior

complex

spike

J. Neurophysiol.

73, 132%1340

of the

one:

information with

-

Attention

involvment

E. and Allen,

neurons:

45 Helmuth.

D.V. and Merzenich, a spatial

network

6, 3555

(1997)

of the cerebellum

discharge

SC;. 19. 501-502

D.V. and Mauk,

cerebellum

role of the cerebellum PET activation

a/.

of motor

43 Courschesne,

Elsevier

al.

Science 267. 1028-1030 G. et

independent

in

et

Psycho/.

realistic

L.L. and Ivry. R.B. (1996)

reduced Hum.

timing

variability

Percept.

When

during

Perform.

two

hands

bimanual

are better

movements

than 1. Exp.

22,278-293

Semantic networks: visualizations of knowledge Roger T. Hartley

and John A. Barnden

n conferences complexity written, b& between

and in the literature,

theory

in computer

their use as a formal

use as an informal

.i.._

science. Many ex

and yet it is our belief that none of them

.(

scheme for knowledge

tool for thinking.

In our

^

T

he history of the development of semantic networks is well known (for an introduction to semantic networks, see

about an entity at a single node and show related information by arcs connected directly to that node”. In contrast, in

Box 1). Both Sowa’ and Lehmann’ have expounded in excellent scholarly fashion as to their origins in the study of

only destroys the readability of the formula, but also obscures

is also well knowr?,

computerized

‘the scattering of information

not

as a tool for representing

the semantic structure of the sentence from which the formula

as is their role in building

was derived. So the battle is joined! The visual aspects of the

language. Their later development knowledge

symbolic logic notations:

inference systemPO.

Indeed, the triad of in-

semantic network

notation

are preferred (at least by Sowa)

telligent thought, logic and language will never be far from our discussion. From all these sources we learn that seman-

over the arcane, but more traditional

tic networks have three main attributes: (1) They originate in the conceptual analysis of language. (2) They can have an expressiveness equivalent to first-

C.S. Peirce before he abandoned it in favor of a diagrammatic form’a.

order logic, at least (although many do not). (3) They can support inference through

an interpreter

that manipulates internal representations.

logic. Interestingly,

this traditional

notation

of symbolic

notion was invented by

In this paper, we hope to show that this argument is only one component of a larger, more complex one involving the nature of semantics; we will also show how different notations can lead to different

systems with different

pragmatic uses.

Many people would go further and say that semantic networks are indistinguishable from formal logic representations”. However, there is something missing here. The visual

Meaning The design and use of a knowledge

aspect of the semantic network

around the business of meaning. Actually,

idea is clearly important.

As

representation

revolves

one spin-off from

Sowa says: ‘network notations are easy for people to read” and this pragmatic aspect ofthe formalism cannot be ignored.

the meaning triangle of Ogden and Richards”.

According

relates objects in the real world, concepts that correspond to

to Sowa: ‘graphs...can keep all the information

Copyright

0 1997, Elsevier

studies in natural language provides a good start, namely,

Science Ltd. All rights

Trends

in

Cognitive

reserved.

1364-6613/97/$17.00

Sciences

-

Vol.

The triangle

PII: 51364-6613(97)01057-7 1.

No.

5,

August

1997