The inadmissibility of Roger Dixon’s affidavit (Part 1)

In this post, we will show:

  • The skills that Roger Dixon said he used are not recognized by Section 212 of the Act
  • Roger Dixon lied when he claimed that Constable Swartz demonstrated to him how Folien 1 was lifted from the DVD holder
  • When Roger Dixon took test lifts from DVD covers he did not follow the procedure that Const Swartz supposedly demonstrated to him
  • Roger Dixon relied on visual comparisons only and he failed to use an assortment of scientific and mathematical techiques to do a proper comparison

What are the skills the Act recognizes?

Section 212(4)(a) of the Act reads as follows:

212    Proof of certain facts by affidavit or certificate

(4)(a) Whenever any fact established by any examination or process requiring any skill-

  • in biology, chemistry, physics, astronomy, geography or geology;
  • in mathematics, applied mathematics or mathematical statistics or in the analysis of statistics;
  • in computer science or in any discipline of engineering;
  • in anatomy or in human behavioural sciences;
  • in biochemistry, in metallurgy, in microscopy, in any branch of pathology or in toxicology; or
  • in ballistics, in the identification of finger prints or palm-prints or in the examination of disputed documents,

is or may become relevant to the issue at criminal proceedings, a document purporting to be an affidavit made by a person who in that affidavit alleges that he or she is in the service of the State or of a provincial administration or is in the service of or is attached to the South African Institute for Medical Research or any university in the Republic or any other body designated by the Minister for the purposes of this subsection by notice in the Gazette, and that he or she has established such fact by means of such an examination or process, shall, upon its mere production at such proceedings be prima facie proof of such fact: Provided that the person who may make such affidavit may, in any case in which skill is required in chemistry, anatomy or pathology, issue a certificate in lieu of such affidavit, in which event the provisions of this paragraph shall mutatis mutandis apply with reference to such certificate.

What skills did Roger Dixon say he used for his experiments?

From Paragraph 5: “... I examined the exhibits, as described in paragraphs 3.1.1 and 3.2.1.1, the drinking glasses, other DVD covers and the lifts produced during the tests, through a process that requires skill in image analysis and comparison.”

“Image analysis and comparison” is not a recognized skill area listed in Section 212 (4)(a).“Image analysis and comparison” is not a recognized skill area listed in Section 212 (4)(a).

This is not a trivial matter that can simply be overlooked and rationalized away.

For example, in Dlamini 2004 (1) SACR 179 (NC) at 180 d – the deponent stated that she had conducted an examination requiring skills in genetics. Because genetics is not one of the fields mentioned in subsections (i) to (iv) of section 212(4)(a) nor was there any information to justify an inference that it forms part of one of the sciences that are mentioned in those provisions the conviction was set aside on review.

Further discussion

Dixon’s investigation consisted of two processes:

Process 1: Devising, and then supervising, experiments that involved taking test lifts with folien from on a number of DVD holders and drinking glasses.

Process 2: Using his skills in image analysis and comparison he analyzed and then compared each test lift with Folien 1.

From Paragraph 5: “... I examined the exhibits, as described in paragraphs 3.1.1 and 3.2.1.1, the drinking glasses, other DVD covers and the lifts produced during the tests, through a process that requires skill in image analysis and comparison.”

Dixon only indicated which skills were required for Process 2 – whereby he compared Folien 1 with each test lift. He, however, did not indicate which skills were required for the first stage where he ‘produced’ each lift.

When a scientist, for example, conducts a DNA analysis or determines the blood alcohol content of a blood sample he/she does so according to specific laboratory protocols and SOPs – according to well-documented industry standards – using calibrated equipment and quality controlled processes. One would therefore expect that the same results would be obtained if other scientists were to analyze the same blood sample. The purpose of a Section 212 Affidavit is then simply to present those results to the court as factual evidence.

Devising an experiment is not a ‘hard science’. In this particular case, if the task were assigned to a different expert, he/she would likely have devised a different experiment/s, which likely would have led to different results and perhaps even in different conclusions than Dixon’s experiment.

There are many variables involved in taking a folien lift from a DVD cover. In addition to the characteristics of the cover itself, there is the technique of the duster, the technique of the lifter, the method of the lifting process, the quality of aluminum powder – to name but a few. When designing an experiment one of the objectives is to control for as many of these variables to the extent possible.

Another expert would likely have involved the duster, Inspector Mariaan Booysens, and the lifter Constable Swartz in an attempt to control for the differences in dusting and lifting techniques, but he chose to use only Captain van Der Westhuizen to perform the dusting and the lifting. Some experts would have used Captain van der Westhuizen and others, in addition to Booysens and Swartz, as a control.

In addition, it is highly likely that another expert would have followed the exact process that Swartz followed when he handled the DVD at the crime scene. According to Swartz, after he received the dusted DVD cover from Booysens, he applied a folien to the front bottom half of the DVD cover. After removing this lift he observed that there were no useable fingerprints on it and he decided to discard this lift. At this stage, the powder over the bottom half has been removed. He then applied a second lift to the top half of the DVD, such that the bottom edge of this second lift overlapped onto the area cleaned by the first lift. This second lift was labeled “Folien #1”.

From Par 6.1 it is clear that Dixon did not follow this procedure:

6.1 Lifts were made of DVD cover #1 as described in paragraph 3.1.1 on the top half of the front of the cover as demonstrated to me by Constable Swartz.

Nowhere in his affidavit is it documented that a first folien was placed over the bottom half before placing a folien over the top half of the front of the cover.

In this Par 6.1 Dixon lied when he claimed that Constable Swartz demonstrated to him how Folien 1 was lifted. The truth is that Constable Swartz has to this day never met Roger Dixon, and has never done any demonstrations to him.

And then in Par 7.5, Dixon concluded:

7.5 ... The concentration of powder on the lower line is not consistent with the line having been produced as a result of a first lift from the middle of the DVD cover, representing the edge of the previous folien, but rather consistent with an edge having been powered.

From Par 7.5 it is clear that Dixon was aware of the “double lift” performed by Swartz, and yet Dixon himself did not perform “double lifts”.       

In conclusion: There are no industry-accepted and documented standards, protocols and methods to determine if a specific folien came from a DVD cover or a drinking glass. The methodology selected by Dixon is just one of many he could have chosen and it is possible that different methodologies could have led to different conclusions. In addition, it seems that there are several very questionable aspects around Dixon’s methodology and skills in devising valid experiments – most critically his failure to follow Swartz’s methodology precisely.

Regarding Stage 2 of his experiment – We previously mentioned that “Image analysis and comparison” is not a recognized skill area listed in Section 212 (4)(a). It would not be accurate to argue that “image analysis and comparison” is a combination of the recognized skill areas e.g. “mathematics” and “identification of finger or palm-prints”.

Dixon’s work with Folien 1 and drinking glasses did not require skills in “identification of finger or palm-prints”. It has already been determined conclusively that the print on Folien #1 belonged to Fred van der Vyver. He was thus not required to identify who the print belonged to.

When investigating whether a certain drinking glass could have made the curves on Folien 1, there are three parameters to consider:

  1. Are the distances between the respective sets of lines the same?
  2. Are the radii of the two respective top curves the same?
  3. Are the radii of the two respective bottom curves the same?

This comparison can be done mathematically – the distance, top radius and bottom radius can be mathematically calculated for a drinking glass of any dimension using basic trigonometry. Then there are various scientific and mathematical techniques to estimate the radii of the existing curves on a folien. It is then possible to scientifically compare the parameters (distance, top radius, bottom radius) as they are on the test lifts with the parameters as they are on Folien 1.

 From Dixon’s affidavit it is evident the only ‘mathematics’ skill he applied was to measure the distance between the curved lines – and even in this, he didn’t do very well.

6.4 The distance between the two parallel curved lines which are visible on Folien 1, as described in paragraph 3.2.1.1, was measured and compared to the heights of the eleven (11) drinking glasses collected from 21 Shiraz flats. Four matching glasses of the eleven were found to have the same height as the distance between the lines ...
Seems Dixon forgot to account for the gap between the bottom of the ruler and ‘0’. The height of this glass is about 83 mm – the average distance between the lines on Folien1 is 80 mm

There is no evidence whatsoever that Dixon mathematically analyzed Folien 1 and the experimental lifts to determine whether the curves were concentric and circular (as they should be if they were made by a conical drinking glass) and also what the radii of the curves are. There is also no evidence that he subjected the dimensions of the different drinking glasses to mathematical analysis to calculate the radii of the curves each glass would have left on a lift.

It appears that the results of Dixon’s comparisons and conclusions are based on a very limited and incomplete mathematical comparison of the key parameters (he only measured and compared distance between lines). It can therefore not be said that he applied his skills in “mathematics”.

It appears that Dixon relied only on a visual comparison.

Leave a comment

Your email address will not be published. Required fields are marked *