(f) The "Second Oswald" Theory Comparison of Oswald Facial Photographs

    Various conspiracy theories have centered around the hypothesis that a double of Lee Harvey Oswald played a part in the assassination of President Kennedy. The theorists themselves appear to disagree on the origin and role of this "Second Oswald." Never agree that in at least one stage of his career between the time Oswald defected to the Soviet Union and the assassination, he was impersonated by a double. (232) To investigate this possibility, the anthropology consultants examined a series of Oswald photographs ranging in time from his Marine Corps enlistment to his arrest in Dallas after the assassination.

    2. ISSUE

  1. Is there any photographic evidence of an Oswald imposter?


  2. The collection of photographs pertaining to the Kennedy assassination and Warren Commission investigation includes several, they vary widely impose, facial expression, lens-subject distance, and image quality. From these, it was possible to select nine in which (1) the facial features were fairly well defined.(2) the pose was either nearly full-face or true profile, and (3) represented the subject. During various key episodes of his life from the time he was a Marine until the assassination. (See figs. IV-60, IV-61, JFK exhibits 556 557.)

    FIGURE IV-60.
    FIGURE IV-61.

    additions, two photographs of definitely poorer quality were selected for analysis. These were two of the controversial "backyard photographs;" they differ from the others in that the direction of lighting was from almost directly overhead and the facial image was somewhat more poorly defined. (See figs. IV-18 and IV--20

  3. In addition to the Oswald photographs, data were included from three photographs of Billy Lovelady, taken in the early 1960's. Lovelady was a fellow employee of Oswald's at the Texas School Book Depository and his strong physical resemblance to Oswald was a ounce of controversy and confusion regarding the "man in the doorway" photograph.* The inclusion of Lovelady's facial indices in our analysis provides a convenient control or yardstick to measure the variation observed in the facial indices derived from the Oswald photographs.

    4. METHODS

  4. This analysis is based on 15 indices derived from 16 measurements of the head and face.** The measurements were taken to the nearest 1 millimeter from 8-by-lO-inch, black-and-white enlargements of the subject's face. The indices for both Oswald and Lovelady are given in table I. There are some missing values for the three profile views of Oswald. This is because certain measurements necessary for calculating these indices cannot, be obtained from a profile photograph also, a few indices could not, be calculated for the full-face photographs because lighting, image clarity, or other factors would not permit the necessary measurements to be made with sufficient accuracy.
    *See par. 759 infra.
    **See addendum A, pars. 74-748 infra.

  5. . In order to reduce this complex set of individual values to more meaningful statistics, one of the methods long employed by anthropologists was used to compare both living and fossil populations. The method selected was Penrose's distance statistic, which has an advantage over more sophisticated multivariate methods in that it is fairly simple to compute, but still gives an acceptable approximation of the morphological differences between the groups. (233) The use of more elaborate methods did not seem justified in view of the small sample sizes involved.

  6. This method reduces it set of complex variables that characterize two or more groups to a pair of coefficients that reflect the groups' overall difference in size and shape.

  7. To apply this method to the present problem, the index data was grouped chronologically to represent Oswald at various significant periods of his life:
    1. Marine Corps.
    2. Russia.
    3. Backyard (Dallas).
    4. New Orleans.
    5. Arrest (Dallas).

  8. The data were then studied to determine whether the face of the individual shown in the Oswald photographs, taken during any one of the first four of these periods (Marine Corps, Russia, backyard, New Orleans), differed morphologically from the face of the man who was arrested in Dallas after the assassination. If such a difference was found, it might suggest that a double was involved.


  9. There are no biological inconsistencies in the Oswald photographs examined that would support the theory that a second person, or double, was involved. The variation observed is that expected in an array of photographs taken by different cameras with varying lens, camera angles, lighting, and other technical differences.

  10. It is not, however, possible totally to dismiss the "second Oswald" hypothesis on the basis of this negative finding. For example, it is possible that ,a double--if one existed--may not have been included in the series of photographs examined. There is also a possibility, however remote, that such a double was such a perfect twin of Oswald that no detectable metric or morphological differences are discernible in the photographic record.


  11. The results of the analysis are shown in the accompanying graph. (See fig. IV-62.) The origin of the graph represents the facial indices of the Dallas arrestee. The various points on the graph represent the other Oswald photographic sets as well as that of Billy Lovelady. The points were determined by plotting the Penrose size coefficient ,against the shape coefficient. The closer a point falls to the origin, the greater the similarity in facial morphology between the individual represented in a particular set of photographs and the person arrested in Dallas. As might be expected, the point representing Billy Lovelady lies much farther from the origin than those representing Oswald. Of the latter, the backyard photographs are the most divergent.

  12. Compared to Lovelady, who strongly resembles Oswald, the Marine, Russia, backyard, and New Orleans photographs cluster rather closely to the origin. It seems highly probable that the relatively small deviation observed in the Oswald data can be, attributed to inevitable error involved in locating landmarks and making measurements from the photographs rather than to the existence of an Oswald double.

  13. In addition to the analysis of facial indices described above, other facial features were compared. For example, in the three profile views, the angle of the nasal bridge in relation to the face was 37(deg in all three cases and the angle between the nasal septum and the facial plane varied by less than 1(deg). The ears are relatively distinctive in shape and are strikingly similar in all photographs where they can be examined. The hairline, if one makes allowance for the passage of time, is quite compatible in all photographs examined. Also, there was no evidence of any incompatible anomalies, scars or other characteristic features suggesting different individuals in the various photographs.

    TABLE III.--Measurements used to derive indices of comparison ofOswald photographs

    1. Total head height__________Vertical distance from the top of the head to the bottom of the chin(vertex to menton N.B.: some allowance for hair must be made in taking this measurement from a photograph.
    2. Physiognomic face height___Distance from the midpoint of the hairline to the most anterior point on the lower forehead just above the nasal root depression (trichion to glabella).
    3. Forehead height____________Distance from the midpoint of the hairline to the most anterior point on the lower forehead just above the nasal root depression (trichion to glabella).
    4. Nose length________________Distance from the deepest point of the nasal root depression to the junction point between the nasal septum and the upper lip(subnasion to subnasale).
    5. Total face height__________Distance between the most anterior point on the helix of ear and the lowermost point on the chin (glabella to menton).
    6. Ear length_________________Distance between the uppermost point on the helix of ear and the lowermost point on the earlobe (superaurale to subaurale).
    7. Lobe length________________Distance between the lowest point in the intertragic notch and lowest point on the earlobe (inter-tragion to subaurale)
    8. Mouth height_______________Distance from the point of contact between the upper and lower lip and lowest point on the chin (stomion to menton).
    9. Maximum head breadth_______Horizontal distance across the broadest portion of the head. N.B.: Some allowance for hair must be made in taking this measurement from a photograph.
    10. Interpalpebral breadth_____Horizontal distance between the inner corners of the eyes (biendocanthion breadth).
    11. Facial breadth_____________Maximum horizontal distance across the face (bizygomatic breadth).
    12. Mandibular breadth_________Horizontal distance between the angles of the jaws (bigonal breadth).
    13. Chin width_________________Horizontal distance across the eminence of the chin.
    14. Chin eminence height_______Distance from the point of deepest depression between the lower lip and chin and the lowest point on the chin (supramentale to menton)
    15. Nasal breadth______________Maximum horizontal breadth across the nasal wings (bialar breadth).
    16. Lateral ear projection_____Horizontal distance from the lateral-most point on the outside margin of the ear to the junction of the ear with the face.

    FIGURE IV-62.--Penrose Size and Shape coefficients Calculated From Facial Indices of Lee Harvey Oswald and Lovelady Photographs. Origin of the Graph Represents Dallas Arrest Photographs.


  14. Table I gives the individual indices calculated from the measurements taken from the photographs of the series under examination. These were averaged for each photographic set to give the mean indices shown in Table II. In some instances these are, of necessity, based on a single value. The mean and standard deviation of the index values of the photographic sets (including Lovelady's) were then computed. These statistics are also given in Table II.

  15. The index values were then converted to standard deviation units (d-values). The Penrose coefficients were calculated from the values, using the following formula:

  16. The size and shape coefficients calculated for the various photographic sets and plotted in Table I are as follows:



  17. Several Warren Commission critics have alleged that substantial differences exist in the reported heights and facial characteristics among different photos and other measurements purported to represent Lee Harvey. Oswald. (234) For example, differences of as much as 2 inches in height exist between an early Marine Corps induction photo.

    Oswald in front of a height chart (see fig. IV-63, JFK exhibit F-166), reported height measurements of Oswald, (235) measurements of the Oswald corpse in Dallas, and another height chart photograph of Oswald (see fig. IV-64). The Marine photograph, which allegedly depicts Oswald with a 13-tach head (measuring from the bottom of his chin to the top of his head), is also said to be inconsistent with his true facial measurements. (236) (See fig. IV-63). On this basis, it has been alleged that these differences are evidence of different individuals purporting to be Lee Harvey Oswald.(237)


  18. Are the differences in Oswald's body measurements, as detected from photographs of him standing against a height chart, probative in any way of an Oswald imposter theory?


  19. Two members of the photographic evidence panel were directed to take an independent series of photographs involving an individual of known height standing against a height chart. For each series of pictures, each person was to be photographed at different distances in relation to the height chart. The vertical orientation of the camera and its distance to the height chart was also subject to change at the photographer's discretion, but the camera was kept essentially horizontal at all times so that optical axis was level, that parallel to the ground.)

  20. In addition, the forensic anthropologists on the photographic evidence panel were asked to provide information concerning discrepancies between measured and reported heights.
    * This section was prepared under the direction of W.K. Hartmann and C.W. Kirk. For related testimony of Kirk, see HSCA-JFK Hearings, vol. IV, pp. 362-65.


  21. No probative weight should be given to an Oswald imposter theory based upon differences in Oswald's body measurements that have been detected from photographs of him standing against a height chart.


  22. First, panel member Dr. William K. Hartmann made a series of photographs of a subject standing with a height chart in the background, but with the subject standing at two different distances from the chart (shoulder-to-chart distance, 1 inch and 10 inches) and the camera at two distances typical of identification camera working disstances (45 inches and 58 inches, respectively, from the height chart). To simulate the typical practice of adjusting the camera to the subject's approximate facial height (sometimes to eye level, sometimes to nose level, et cetera), the camera was elevated and lowered through a series of different vertical positions from chin level to the top of the hair. It was found that the subject's height, read from the height chart, ranged from 0 to 1 3/4 inches higher than the actual measured hair-top position during these photos. (238) *

  23. A second test involving similar procedures was conducted by Sergeant Cecil W. Kirk, of the District of Columbia Metropolitan Police Department's mobile crime lab, using departmental identifica- tion camera and height chart equipment. The subject's height in this experiment, as read from the height chart, increased one-half inch as the subject moved from a position with heels against the wall to a position with heels 8 inches out from the wall. In addition, while the height of the subject's head actually measured 8 inches from chin to head top, the readings on the height chart were approximately 12 1/2 and 14 inches, thereby resulting in errors of 4 1/2 and 6 inches, respectively. (230) (See fig. IV-65, JFK exhibit F-564.)

  24. The types of discrepancies obtained in the Hartmann and Kirk studies are attributable to parallax errors which, in this case, present a difference in scale between the images of the subject and the chart. Parallax errors occur because the plane of the subject's face or body is not in the same plane as the height chart to which it is being compared since these two planes were photographed from a finite distance, nonparallel lines from camera to subject were introduced. The nonparallel lines diverge from the camera, lens to the subject. Consequently, from a camera centered in front of the subject's face, the line of sight from the camera lens slopes upward past the top of the subject's head, yielding a higher reading on the background wall chart than the actual head-top height.

  25. Moreover, unless the subject photographed is standing with his back against the height chart at a correct distance from a properly positioned camera equipped with an appropriate lens,* it is unreasonable to assume that the resulting picture is ever a precisely accurate indicator of both his height and head size. (240) For this reason, height charts are no longer commonly used in law enforcement and industrial security work. (241)
    *In addition, because this particular subject's driver's license reported his height as 1 3/4 inches smaller than his actual measured hair-top position during the photography, the total discrepancy between the height chart readings and the driver's license ranged from 1 3/4 to 3 1/2 inches. The reported height in the subject's driver's license was not checked until after the photography had been completed.

  26. Finally, the photographic evidence panel's board of forensic anthropologists advised that a diurnal variation in height of half an inch or more is common during the course of a day, with the subject generally being taller in the morning when the spine has been less compressed. (242) The board also cited to the panel an anthropological study by Robert M. White and Edmund Churchill ("The Body Size of Soldiers," U.S. Army Natic Laboratories, technical report 72-51-CE, 1971), which measured heights of 6,682 army personnel versus the heights these individuals reported for themselves. Typical discrepancies in ]might were 1.1 inches. Generally, men of average height (5 feet 9 inches) reported themselves 1.1 inches taller than their measured stature; relatively short, men reported themselves about 0.8 inch taller; and relatively tall men reported themselves 1.2 inches taller.
    * To get an accurate height measurement, the camera must be level and its optical axis must be level with the top of the head.

    FIGURE IV-63
    FIGURE IV-64.--New Orleans arrest picture of Oswald in front of height chart.
    FIGURE IV-65.

g. Comparison of photographs of Lee Harvey Oswald and Billy Nolan Lovelady with that of a motorcade spectator