- Email: [email protected]
Contrast sensitivity function of the infant visual system
CONTRAST
SENSITIVITY FUNCTION INFANT VISUAL SYSTEM’
MMTIX S. B~NRS and
PHILIP
OF THE
SALAPATW
Learning. Institute of Child Development and Center for Research - _- in_AHuman Cniversity of kfinnesota. Minneapolis, >fh >>4>3. U.S.A.
Abstract-The response of the human infant visual system to sinewave gratings of various spatial frequencies was measured. The contrast sensitivity functions obtained are an e&mate of the spatial information available to the infant. Evidence for lateral inhibitory processing was found. The implications of the results For the development of form perception are discussed. ISTRODLCTION
field. Contrast sensitivity, however, was considerably lower than that of an adult. It should be noted that the use of flashing or drifting gratings yields CSFs in adults of consi~~bly different form than CSFs determined with stationary grating (Robson, 1966; van Nes, Koenderink, Nas and Bouman, 1967). This difference is most pronounced for spatial frequencies less than 6 cyclesideg which is the approximate range of the infant CSF reported. Thus, the Atkinson ef al. (1974) data provide little ~fo~ation concerning form vision for stationary targets. We have measured the CSF of five Z-month-old infants using stationary sinewave grat-
one might take in studying the development of form vision. One, which has been popular in the last decade, is to present a wide variety of forms to infants and determine which they tend to fixate more frequently. If a clear fixation pre-
There
are several approaches
ference appears for a given form, it is inferred that infants can discriminate this form from ot_hers(Fantt, 1958; Kessen, Haith and Salapatek, 1970). Using this approach, researchers have shown that infants can discriminate forms on a variety of stimulus attributes. This approach, however, has not been particularly successful in uncovering general properties of infant form \rision since it does not readily lend itself to generalization beyond the particular forms presented. A second approach with perhaps greater generality has been used successfully in the study of adult form vision: the measurem~t of the contrast sensitivity function (CSF). The CSF is determined by measuring an observer’s contrast threshold for sinewave gratings of various spatial frequencies. Since any visual stimulus can be represented by the addition of a limited set of sinewave gratings, the CSF provides considerable isolation about form vision in general (Comsweet. 1970). Measurement of a Z-month-old infant’s CSF has recently been reported by Atkinson, Braddick and Braddick (1974). Flashing and drifting sinewave gratings were used in a fixation preference paradigm. The results showed that the infant could discriminate OX--Scycles,ideg gratings from an even luminance ’ Research supported by NICHD grant No. HD-IX098 to the Center for Research in Human Learning, NICHD grant No. HD-05027 to the Institute of Child Development. and by NIH grant No. ROl HD07317 to Phitip Sala. Patek. We thank E. P&OS, J. Simonson. R. hfonson for assistance in data coIIection, D. E. Mitchell for commits on an earlier draft, and K. Adams for secretarial assistance. ’ The four sides of the rotating mirror (M2) were within 0.005 in. of true parallel about the axis of rotation. Furthermore, the orientation of T, MI and the axis of Zvl, were finely adjustable. Thus, with suitable adjustment of the apparatus. no significant reduction in contrast between 0.15 and 1 cyclesideg (the range of spatial frequencies used) was observed when the stimuli were measured photometriGlfIy.
in@.
METHODS The apparatus is schematized in Fig. I. Gratings were produced optically in a manner similar to that described by Da\-idson ( 1968). Transparencies were prepared with a sinusoidal profile over half of their horizontal extent and a straight-line profile over the other half. The transparencies were opaque below the profile and transparent above it. A given transparency was imaged at the plane of the opa[-flashed glass stimulus Reid (F) by the objective lens (Lt). front-surface mirror (M,), and a four-sided rotating mirror (M2)2. The image was thus swept vertically across the stimulus field (F) at a rate of 120 Hz to produce a sinewave gating across one half of the stimulus field and an immediately adjacent uniform field across the other half. Contrast and spatial frequency were varied by inserting different transparencies in the apparatus. Contrast varied with the peak-to-trough ampfitude of the sinewave portion of the transparencies. Spatial frequency varied with the number of cycles in the sinewave portion. The spaceaverage luminance of the grating half-field and the uniform half-field was 16 ft-L. The stimulus field was very large (96’ x 40’) to maximize the infants’ attention to the stimulus tield and to insure a sufficient number of cycles for lower spatial frequencies. The field was reflected by a large front-surface mirror (M,) so that the infants could view the stimuli while lying in a supine position. The optical tf;;e between the infants and the stimulus field was
A &ond channel (not shown in Fig. 1) imaged a vertical bar at the midline of the stimulus field between presentations of the gatings. When an observer judged that the infant was fixatmg the bar, a 3-set pre~ntation of the grating and uniform half-fields occurred. Gratings of variable 867
spatial frcqusncq and contrast uere presented in quasi-random order to either the left or right of center. Two obser\ers. uho could not see the stimulus field. indicated whether the infants’ first fixation was to the left or right. The obssrwzrs’ responses were automatically registered on an twnt recorder. The experimenter determined the number of hits and misses on-line by comparing the obscrlers‘ responses to the side on which the grating had been presented. A hit was recorded uhen both observers indicated that the first fixation was to the side the gating had been on. A miss was recorded when both observers indicated that the first fixation was to the side the uniform field had been on. When the observers’ responses were contradictory ione indicated the first fixation was to the left and the other observer indicated it WASto the tight), the experimenter used a random array to assign a hit or miss to the trial. When the observers did not indicate that an eye mosement had occurred. a simiiar random assignment procedure was used. This did not occur very often since the observers were encouraged to guess when they were uncertain of the direction of first fxxation. A descendingstaircase procedure was employed to approximate a given infant’s contrast threshold. The procedure continued until a contrast associated with a hit rate below 7Y0 and a contrast associated with a rate above X0, had been presented. A total of 10 trials at each of these contrasts was presented. The final estimate of contrast threshold was obtained by interpolation co the contrast associated with a hit rate of cvactl~ 75”,. We presented a total of 2#-300 trials to each infant. RESL;L-iS ASD DISCUSSION The CSFs obtained for the five infants are shown in Fig. 7. The agreement among these functions is
quite good for infant data. All of the infant CSFs exhibit rn~~urn sensitivity to mid-Frequency gratings and lower sensitivity to both Iow-(0.15 cycles/deg) and highfrequency (1 and 2 cyclesjdeg) gratings. Three of the infants showed peak sensitivity to 0.3 cycfesjdeg, and the other two to 0.5 cyclesjdeg. Figure 3 shows the average of the five infant CSFs 22
~_-.‘---
~-
--.-.
-
-..
I
-:
2. Contrast sensitivity functions (CSFs) for each of the five Z-month-old infants. Contrast threshold is plotted against spatial frequency. Contrast is defined as: (L,,,,,,- L,,,i, ),(&,,U,+ L,& where I!_,,,,,is the luminance of the most intense part of the grating and L,,,, the luminance of the least intense part. NT designates no threshold obtained. Fig.
Fig. >. Average CSF for the Z-month-old infants and the CSF for an adult observer. Contrast threshold is plotted against spatial frequency. The dashed lmz represents 3 typical high-frequency portion for an adult CSF (tan >lteteren. 1967).
and the CSF obtained for an adult observer (who viewed the stimuli while fixating a central fisation p&i%). The highest spatial frequency an adult observer can detect is a common measure of acuity ~Campbell and
Green, 1965). This spatial frequency can be estimated by linearly extrapolating the high-frequency slope of the CSF when it is plotted in semilog coordinates. Extrapolation of the high-frequency slope of the a~age CSF (best-fitting least-squares hne through rhe thresholds at 0.5, 1 and 2 cyclesidegr reported here yields an estimate of 1.1 cycleudeg as the highest spatial frequency detectable by 2-month-olds. in Snellen notation this corresponds to an acuity of 20.250, a value higher than most previous reports of stationary acuity (Fantz Ordy and Udelf, 1962: Teller, Morse. Borton and Regal, 1971: Saiapatek. Bechtold and Bushnell, 19753. but lower than reports of non-srationary acuity (Atkinson er 01.. 1974: Dayton. Jones. Aiu Rawson. Steele and Rose, 1964). Of more interest is the low-frequency decrease in sensitivity. This low-frequency turndown is commonly observ.ed in adult CSFs (Rathff. 1965: Schade. 1956). Until recently, the turn-down was assumed to result from iateral inhibitory processing in the visual system (Ratliff, 1965; Made, 1956). TWO groups of investigators have questioned this assumption arguing that detection of a low-frequency grating is determined by its totai number of cycles and not by its spatial frequency (if the total number of cycles is less than 7) (Savoy and McCann, 1975: Hoekstra tir (il.. 1971). To avoid possible confounding of number of cycles and spatial frequency, we presented stimuli in a wide field. The number of cycles was seven or greater for ail spatial frequencies presented. Our adult data show that when the stimulus tirld is widened and a fixation point provided, the low&equency turndown is still observed (Fig. 3). Under these conditions. low-frequent)-: ~sensitivit~ is probably related to lateral inhibitron. Thus, the infant CSFs reported here suggest that spatial information undergoss some form of lateral inhibitory processing in the young infant‘s IkXld
system.
Fig. 1. Schematic diagram of the apparatus. LS, light sowce. iLi and Lr, fenses. T, transparency. S, shutter. M1 and MS, front-surface mirrors. W+ rotating four-sided mirror. A, aperature. F, flashed opal glass stimulus field. Also shown are a typical transparency and the stimulus pattern produced by that transparency,
[_fmi~g page 868
Contrast sensitivity function of the infant visual system
869
REFERESCES Atkinson tq td. (19741 found only weak ev.idence for a low-frequency turn-down in the infant CSF and Atkinson J.. Braddick 0. and Braddick F. (1974) Acuity concluded that “the visual pathway of the infant may and contrast sensitivity of infant vision. S~trnre, Lontl. be so organized as to transmit low spatial frequencies 211, u)3304. relatively well’ (p. Uw) due to an underdevelop Campbell F. W. and Green D. G. (1965) Optical and state of laterai inhibitory processing. There are wo retinal. factors affecting visual resolution. J. Ph.rsiol.. Lo&. 181, 576-59;. ways in which the discrepancy between our results and those of Atkinson er al. might reasonably be Comsweet T. N. (1970) C’isuctlPerception Academic Press, New York. accounted for. First, it has been shown that the use of flashing or drifting gratings tends to enhance Iow- Davidson M. (1968) Pert~rbatjon approach to spatial brightness interaction in human vision. J. opt. Sot. rtm. frequency sensitivity 111adults (Robson, 1966; van Nes 58, 1300-1308. et af., 1967). If this is.true in infants as well, Atkinson Dayton G.. Jones M.. Aiu P., Rawson R.. Steele B. and et al. may have obscured the low-frequency turnRose M. i 1964 Derelopmental study of coordinated eye down by temporally m~uIat~g their gratings. This moYernew m the human infant-i: Visual acuity in the criticism is somewhat weakened by the fact that the newborn human: a study based on induced optokinetic modulation rates they used were low enough to nysta_gmus recorded by electrooculography. -4rchs Ophcl&. 71, 565-870. observe a low-frequency turn-down in their adult observer. Second, the lowest spatial frequency pre- Fantz R. L. (1958) Pattern vision in young infants. Ps.whol. Rrc 8, 6-47. sented in the Atkinson er al. study was 0.X cycles:deg. Fantz R. L.. Ordv J. and Udelf kf. (1962) Maturation of In the present study three of the infants exhibited pattern vision in infants during the first 6 months of low-frequency insensitivity to only 0.15 cycles.‘deg. life. J. camp. plluiof. Psyhol. 55, 907-917. Thus, it is possible that spatial frequencies lower than Hoekstra J.. -van der Go& D. P. J.. van den Brink G. those employed by Atkinson es ~1. are required to and Bilsen F. A. 11974) The inffuence of the number demonstrate the low-frequency turn-down in infants. of cycles upon the’ visual contrast threshold for spatial The presence of lateral inhibitory processing has sine-wave patterns. C’ision Res. 14, 36%368. important implications for the understanding of in- Kessen W., Haith 11. M. and Salapatek P. (1970) Human infancy: a bibliography and guide. fn C~rmichueTs fant form vision. By making the visual system rela.~~~~~~~~~f of Child ~~~~~1~~~~~~~ IEdtted by Mussen P.), Vol. tively insensitive to low spatial frequencie_s, lateral in1. Wiley. New York. hibition serves to increase the Gsibitity of sharp Qunnritnrirr Sclrdies on changes in illumination (edges) while decreasing the Ratliff F. (1965) JIcrch Bnntls: Newtrl NerworX-s in the Retina. Holden-Day. San Franvisibility of gradual changes in illumination. This in cisco. turn is thought to facilitate the process of form per- Robson J. G. (1966) Spatial and temporal contrast-sensitiception by enhancing the visibility of the contours vity functions of the visual svstem. J. oar. Sac. Am. 56. of forms (Cornsweet. 1970; Ratliff. 1965). Our data llil-114’. suggest that this mechanism for emphasizing edge in- Sulapatek P.. Bechtold A. G. and Bushnell E. W. (1975) formation is present at or shortly after birth. Infant visual acuity as a function of viewing distance. Clritd Der. (in press). When one compares the infant and adult CSFs Savoy R. L. and McCann J. J. (i973) Visibility of low shown in Fig. 3, it is apparent that a comparatively spatial frequency sine-wave targets: dependence on very limited range of spatia1 information is available number of cycles. J. opt. Sot. .-lm. 65. 3J_Z-350. to the infant. Thus, despite their ability to discrimiSchade 0. H. (1956) Optical and photoelectric analog of nate some forms (Fantz, 1958; Kessen, Haith and the eye. J. opt. Sue. ilm. 46, 7X-739. Salapatek, f970), the visual world of young infants Teller 0. Y., Morse R., Borton R. and Regal D. (1971) is quite impoverished compared to our own. The Visual acuity for vertical and diagonal gratings in CSFs reported here give a quantitative estimate of human infants. Vision Res. 11, 1133-1439. how great this impoverishment is (at least for stationvan Meeteren A. (1967) Spatial Sinewave Response of the ary stimuli). Furthermore. to the extent that the infant Visual System. Inst. for Perception TNd. Rept No. IZF-1966-7. visual system is a linear system (an assumption not van Nes F. L.. Koenderink J. J.. Nas H. and Bouman yet tested), the CSF can be used to estimate the M. A. (1967) Spatiotemporal modulation transfer in the appearance of simple two-dimensional visual stimuli human eye. J. opr. Sot. ;lm. 57, 1082-1088. to the infant.
SENSITIVITY FUNCTION INFANT VISUAL SYSTEM’
MMTIX S. B~NRS and
PHILIP
OF THE
SALAPATW
Learning. Institute of Child Development and Center for Research - _- in_AHuman Cniversity of kfinnesota. Minneapolis, >fh >>4>3. U.S.A.
Abstract-The response of the human infant visual system to sinewave gratings of various spatial frequencies was measured. The contrast sensitivity functions obtained are an e&mate of the spatial information available to the infant. Evidence for lateral inhibitory processing was found. The implications of the results For the development of form perception are discussed. ISTRODLCTION
field. Contrast sensitivity, however, was considerably lower than that of an adult. It should be noted that the use of flashing or drifting gratings yields CSFs in adults of consi~~bly different form than CSFs determined with stationary grating (Robson, 1966; van Nes, Koenderink, Nas and Bouman, 1967). This difference is most pronounced for spatial frequencies less than 6 cyclesideg which is the approximate range of the infant CSF reported. Thus, the Atkinson ef al. (1974) data provide little ~fo~ation concerning form vision for stationary targets. We have measured the CSF of five Z-month-old infants using stationary sinewave grat-
one might take in studying the development of form vision. One, which has been popular in the last decade, is to present a wide variety of forms to infants and determine which they tend to fixate more frequently. If a clear fixation pre-
There
are several approaches
ference appears for a given form, it is inferred that infants can discriminate this form from ot_hers(Fantt, 1958; Kessen, Haith and Salapatek, 1970). Using this approach, researchers have shown that infants can discriminate forms on a variety of stimulus attributes. This approach, however, has not been particularly successful in uncovering general properties of infant form \rision since it does not readily lend itself to generalization beyond the particular forms presented. A second approach with perhaps greater generality has been used successfully in the study of adult form vision: the measurem~t of the contrast sensitivity function (CSF). The CSF is determined by measuring an observer’s contrast threshold for sinewave gratings of various spatial frequencies. Since any visual stimulus can be represented by the addition of a limited set of sinewave gratings, the CSF provides considerable isolation about form vision in general (Comsweet. 1970). Measurement of a Z-month-old infant’s CSF has recently been reported by Atkinson, Braddick and Braddick (1974). Flashing and drifting sinewave gratings were used in a fixation preference paradigm. The results showed that the infant could discriminate OX--Scycles,ideg gratings from an even luminance ’ Research supported by NICHD grant No. HD-IX098 to the Center for Research in Human Learning, NICHD grant No. HD-05027 to the Institute of Child Development. and by NIH grant No. ROl HD07317 to Phitip Sala. Patek. We thank E. P&OS, J. Simonson. R. hfonson for assistance in data coIIection, D. E. Mitchell for commits on an earlier draft, and K. Adams for secretarial assistance. ’ The four sides of the rotating mirror (M2) were within 0.005 in. of true parallel about the axis of rotation. Furthermore, the orientation of T, MI and the axis of Zvl, were finely adjustable. Thus, with suitable adjustment of the apparatus. no significant reduction in contrast between 0.15 and 1 cyclesideg (the range of spatial frequencies used) was observed when the stimuli were measured photometriGlfIy.
in@.
METHODS The apparatus is schematized in Fig. I. Gratings were produced optically in a manner similar to that described by Da\-idson ( 1968). Transparencies were prepared with a sinusoidal profile over half of their horizontal extent and a straight-line profile over the other half. The transparencies were opaque below the profile and transparent above it. A given transparency was imaged at the plane of the opa[-flashed glass stimulus Reid (F) by the objective lens (Lt). front-surface mirror (M,), and a four-sided rotating mirror (M2)2. The image was thus swept vertically across the stimulus field (F) at a rate of 120 Hz to produce a sinewave gating across one half of the stimulus field and an immediately adjacent uniform field across the other half. Contrast and spatial frequency were varied by inserting different transparencies in the apparatus. Contrast varied with the peak-to-trough ampfitude of the sinewave portion of the transparencies. Spatial frequency varied with the number of cycles in the sinewave portion. The spaceaverage luminance of the grating half-field and the uniform half-field was 16 ft-L. The stimulus field was very large (96’ x 40’) to maximize the infants’ attention to the stimulus tield and to insure a sufficient number of cycles for lower spatial frequencies. The field was reflected by a large front-surface mirror (M,) so that the infants could view the stimuli while lying in a supine position. The optical tf;;e between the infants and the stimulus field was
A &ond channel (not shown in Fig. 1) imaged a vertical bar at the midline of the stimulus field between presentations of the gatings. When an observer judged that the infant was fixatmg the bar, a 3-set pre~ntation of the grating and uniform half-fields occurred. Gratings of variable 867
spatial frcqusncq and contrast uere presented in quasi-random order to either the left or right of center. Two obser\ers. uho could not see the stimulus field. indicated whether the infants’ first fixation was to the left or right. The obssrwzrs’ responses were automatically registered on an twnt recorder. The experimenter determined the number of hits and misses on-line by comparing the obscrlers‘ responses to the side on which the grating had been presented. A hit was recorded uhen both observers indicated that the first fixation was to the side the gating had been on. A miss was recorded when both observers indicated that the first fixation was to the side the uniform field had been on. When the observers’ responses were contradictory ione indicated the first fixation was to the left and the other observer indicated it WASto the tight), the experimenter used a random array to assign a hit or miss to the trial. When the observers did not indicate that an eye mosement had occurred. a simiiar random assignment procedure was used. This did not occur very often since the observers were encouraged to guess when they were uncertain of the direction of first fxxation. A descendingstaircase procedure was employed to approximate a given infant’s contrast threshold. The procedure continued until a contrast associated with a hit rate below 7Y0 and a contrast associated with a rate above X0, had been presented. A total of 10 trials at each of these contrasts was presented. The final estimate of contrast threshold was obtained by interpolation co the contrast associated with a hit rate of cvactl~ 75”,. We presented a total of 2#-300 trials to each infant. RESL;L-iS ASD DISCUSSION The CSFs obtained for the five infants are shown in Fig. 7. The agreement among these functions is
quite good for infant data. All of the infant CSFs exhibit rn~~urn sensitivity to mid-Frequency gratings and lower sensitivity to both Iow-(0.15 cycles/deg) and highfrequency (1 and 2 cyclesjdeg) gratings. Three of the infants showed peak sensitivity to 0.3 cycfesjdeg, and the other two to 0.5 cyclesjdeg. Figure 3 shows the average of the five infant CSFs 22
~_-.‘---
~-
--.-.
-
-..
I
-:
2. Contrast sensitivity functions (CSFs) for each of the five Z-month-old infants. Contrast threshold is plotted against spatial frequency. Contrast is defined as: (L,,,,,,- L,,,i, ),(&,,U,+ L,& where I!_,,,,,is the luminance of the most intense part of the grating and L,,,, the luminance of the least intense part. NT designates no threshold obtained. Fig.
Fig. >. Average CSF for the Z-month-old infants and the CSF for an adult observer. Contrast threshold is plotted against spatial frequency. The dashed lmz represents 3 typical high-frequency portion for an adult CSF (tan >lteteren. 1967).
and the CSF obtained for an adult observer (who viewed the stimuli while fixating a central fisation p&i%). The highest spatial frequency an adult observer can detect is a common measure of acuity ~Campbell and
Green, 1965). This spatial frequency can be estimated by linearly extrapolating the high-frequency slope of the CSF when it is plotted in semilog coordinates. Extrapolation of the high-frequency slope of the a~age CSF (best-fitting least-squares hne through rhe thresholds at 0.5, 1 and 2 cyclesidegr reported here yields an estimate of 1.1 cycleudeg as the highest spatial frequency detectable by 2-month-olds. in Snellen notation this corresponds to an acuity of 20.250, a value higher than most previous reports of stationary acuity (Fantz Ordy and Udelf, 1962: Teller, Morse. Borton and Regal, 1971: Saiapatek. Bechtold and Bushnell, 19753. but lower than reports of non-srationary acuity (Atkinson er 01.. 1974: Dayton. Jones. Aiu Rawson. Steele and Rose, 1964). Of more interest is the low-frequency decrease in sensitivity. This low-frequency turndown is commonly observ.ed in adult CSFs (Rathff. 1965: Schade. 1956). Until recently, the turn-down was assumed to result from iateral inhibitory processing in the visual system (Ratliff, 1965; Made, 1956). TWO groups of investigators have questioned this assumption arguing that detection of a low-frequency grating is determined by its totai number of cycles and not by its spatial frequency (if the total number of cycles is less than 7) (Savoy and McCann, 1975: Hoekstra tir (il.. 1971). To avoid possible confounding of number of cycles and spatial frequency, we presented stimuli in a wide field. The number of cycles was seven or greater for ail spatial frequencies presented. Our adult data show that when the stimulus tirld is widened and a fixation point provided, the low&equency turndown is still observed (Fig. 3). Under these conditions. low-frequent)-: ~sensitivit~ is probably related to lateral inhibitron. Thus, the infant CSFs reported here suggest that spatial information undergoss some form of lateral inhibitory processing in the young infant‘s IkXld
system.
Fig. 1. Schematic diagram of the apparatus. LS, light sowce. iLi and Lr, fenses. T, transparency. S, shutter. M1 and MS, front-surface mirrors. W+ rotating four-sided mirror. A, aperature. F, flashed opal glass stimulus field. Also shown are a typical transparency and the stimulus pattern produced by that transparency,
[_fmi~g page 868
Contrast sensitivity function of the infant visual system
869
REFERESCES Atkinson tq td. (19741 found only weak ev.idence for a low-frequency turn-down in the infant CSF and Atkinson J.. Braddick 0. and Braddick F. (1974) Acuity concluded that “the visual pathway of the infant may and contrast sensitivity of infant vision. S~trnre, Lontl. be so organized as to transmit low spatial frequencies 211, u)3304. relatively well’ (p. Uw) due to an underdevelop Campbell F. W. and Green D. G. (1965) Optical and state of laterai inhibitory processing. There are wo retinal. factors affecting visual resolution. J. Ph.rsiol.. Lo&. 181, 576-59;. ways in which the discrepancy between our results and those of Atkinson er al. might reasonably be Comsweet T. N. (1970) C’isuctlPerception Academic Press, New York. accounted for. First, it has been shown that the use of flashing or drifting gratings tends to enhance Iow- Davidson M. (1968) Pert~rbatjon approach to spatial brightness interaction in human vision. J. opt. Sot. rtm. frequency sensitivity 111adults (Robson, 1966; van Nes 58, 1300-1308. et af., 1967). If this is.true in infants as well, Atkinson Dayton G.. Jones M.. Aiu P., Rawson R.. Steele B. and et al. may have obscured the low-frequency turnRose M. i 1964 Derelopmental study of coordinated eye down by temporally m~uIat~g their gratings. This moYernew m the human infant-i: Visual acuity in the criticism is somewhat weakened by the fact that the newborn human: a study based on induced optokinetic modulation rates they used were low enough to nysta_gmus recorded by electrooculography. -4rchs Ophcl&. 71, 565-870. observe a low-frequency turn-down in their adult observer. Second, the lowest spatial frequency pre- Fantz R. L. (1958) Pattern vision in young infants. Ps.whol. Rrc 8, 6-47. sented in the Atkinson er al. study was 0.X cycles:deg. Fantz R. L.. Ordv J. and Udelf kf. (1962) Maturation of In the present study three of the infants exhibited pattern vision in infants during the first 6 months of low-frequency insensitivity to only 0.15 cycles.‘deg. life. J. camp. plluiof. Psyhol. 55, 907-917. Thus, it is possible that spatial frequencies lower than Hoekstra J.. -van der Go& D. P. J.. van den Brink G. those employed by Atkinson es ~1. are required to and Bilsen F. A. 11974) The inffuence of the number demonstrate the low-frequency turn-down in infants. of cycles upon the’ visual contrast threshold for spatial The presence of lateral inhibitory processing has sine-wave patterns. C’ision Res. 14, 36%368. important implications for the understanding of in- Kessen W., Haith 11. M. and Salapatek P. (1970) Human infancy: a bibliography and guide. fn C~rmichueTs fant form vision. By making the visual system rela.~~~~~~~~~f of Child ~~~~~1~~~~~~~ IEdtted by Mussen P.), Vol. tively insensitive to low spatial frequencie_s, lateral in1. Wiley. New York. hibition serves to increase the Gsibitity of sharp Qunnritnrirr Sclrdies on changes in illumination (edges) while decreasing the Ratliff F. (1965) JIcrch Bnntls: Newtrl NerworX-s in the Retina. Holden-Day. San Franvisibility of gradual changes in illumination. This in cisco. turn is thought to facilitate the process of form per- Robson J. G. (1966) Spatial and temporal contrast-sensitiception by enhancing the visibility of the contours vity functions of the visual svstem. J. oar. Sac. Am. 56. of forms (Cornsweet. 1970; Ratliff. 1965). Our data llil-114’. suggest that this mechanism for emphasizing edge in- Sulapatek P.. Bechtold A. G. and Bushnell E. W. (1975) formation is present at or shortly after birth. Infant visual acuity as a function of viewing distance. Clritd Der. (in press). When one compares the infant and adult CSFs Savoy R. L. and McCann J. J. (i973) Visibility of low shown in Fig. 3, it is apparent that a comparatively spatial frequency sine-wave targets: dependence on very limited range of spatia1 information is available number of cycles. J. opt. Sot. .-lm. 65. 3J_Z-350. to the infant. Thus, despite their ability to discrimiSchade 0. H. (1956) Optical and photoelectric analog of nate some forms (Fantz, 1958; Kessen, Haith and the eye. J. opt. Sue. ilm. 46, 7X-739. Salapatek, f970), the visual world of young infants Teller 0. Y., Morse R., Borton R. and Regal D. (1971) is quite impoverished compared to our own. The Visual acuity for vertical and diagonal gratings in CSFs reported here give a quantitative estimate of human infants. Vision Res. 11, 1133-1439. how great this impoverishment is (at least for stationvan Meeteren A. (1967) Spatial Sinewave Response of the ary stimuli). Furthermore. to the extent that the infant Visual System. Inst. for Perception TNd. Rept No. IZF-1966-7. visual system is a linear system (an assumption not van Nes F. L.. Koenderink J. J.. Nas H. and Bouman yet tested), the CSF can be used to estimate the M. A. (1967) Spatiotemporal modulation transfer in the appearance of simple two-dimensional visual stimuli human eye. J. opr. Sot. ;lm. 57, 1082-1088. to the infant.