Timothy Bornyk was charged with break and enter. The police had discovered a fingerprint in the study of a home [latent print] where there had been a break and enter. An expert in fingerprint identification and comparison testified that the print matched Bornyk’s prints [known print]. Notwithstanding this evidence, the trial judge acquitted Bornyk: 2013 BCSC 1927. In doing so the trial judge, inter alia, did his own research, reviewing and then relying upon, some academic articles on fingerprints analysis and did his own comparison of the latent and known prints. The Crown appealed. The British Columbia Court of Appeal allowed that appeal: 2015 BCCA 28.

The issues that were discussed on appeal began at the conclusion of the trial. The trial judge sent counsel a memorandum listing four articles which were critical of fingerprint identification analysis and asked counsel to make further submissions. Crown counsel subsequently sent three additional articles to the trial judge. Oral submissions were later offered by both Crown and defence.

With respect to the articles, the Crown argued that the issues raised therein had no application to the evidence in the present case. The Crown also argued that the articles were not properly evidence and should not be considered [@3].

The Crown also opposed the suggestion that the trial judge could assess the known and latent prints himself and identify differences that undermine the identification made by the expert witness.

The expert was never recalled. The articles were not marked as exhibits.

The trial judge acquitted Bornyk. In doing so he referred to and quoted certain portions of the articles. He further undertook his own comparison of the prints, based on submissions of defence:

In argument, defence counsel noted unexplained discrepancies between the latent and the known fingerprints. Of particular note, in the area of the latent fingerprint stated to be of “low tolerance” and “extremely reliable”, two gaps on the latent fingerprint are not visible on the known fingerprint.

If one goes to the ridge immediately to the left of the respective red dots marking the centre of the delta on the latent and the known fingerprints and traces a line towards the top of the page, on the known fingerprint there is a continuous ridge, whereas on the latent fingerprint there is a gap, a further ridge, another gap, and then a further ridge. [Emphasis added]; [@55-56].

These two approaches, the Crown said, were wrong in law. The Court of Appeal agreed.

With respect to referencing the articles, the Court began by noting that it is “basic to trial work that a judge may only rely upon the evidence presented at trial, except where judicial notice may be taken” [@8]: see R v RSM, 1999 BCCA 218 at para 20; R v Cloutier, 2011 ONCA 484.

In this case the articles, which offered opinions on fingerprint analysis, were “not matters of which the judge could take judicial notice. It is axiomatic that it was not open to the judge to embark on his independent investigation” [@10]. The Court continued:

By his actions, the judge stepped beyond his proper neutral role and into the fray. In doing so, he compromised the appearance of judicial independence essential to a fair trial. While he sought submissions on the material he had located, by the very act of his self-directed research, in the words of Justice Doherty in R. v. Hamilton (2004), 2004 CanLII 5549 (ON CA), 189 O.A.C. 90, 241 D.L.R. (4th) 490 at para. 71, he assumed the multi-faceted role of “advocate, witness and judge”. [@11].

The trial judge further erred when he considered the fruits of his investigation.

Not the least of the problems with the approach adopted by the judge is it opened the door to a mistaken comprehension and application of the information in the articles even if in the field of fingerprint analysis they would be considered authoritative and applicable to procedures employed in this case, an assumption not established in the evidence.

[…]

It is clear from the reasons for judgment that the articles had a material bearing on the acquittal as the judge relied upon them to find that the fingerprint identification was not reliable. Most of the “troubling aspects” he identified were not put to the expert witness, and appear to respond to the articles he located. [@14 and 16]

With respect to the comparison of the fingerprints done by the trial judge, the court found that to be in error as well.

The very point of having an expert witness in a technical area, here fingerprint analysis, is that the specialized field requires elucidation in order for the court to form a correct judgment: Kelliher (Village) v. Smith, 1931 CanLII 1 (SCC), [1931] S.C.R. 672; R. v. Mohan, 1994 CanLII 80 (SCC), [1994] 2 S.C.R. 9, 114 D.L.R. (4th) 419. While it may be desirable that a judge personally observe the similarities and differences between the latent point and known point, such examination should be guided by a witness so as to avoid the trier of fact forming a view contrary to an explanation that may be available if only the chance were provided to proffer it.

The judge relied upon his own observation of what he said was a difference between the latent and known prints. The fingerprint witness however was never questioned on that area of the fingerprint. Whether this “difference” is forensically significant is speculation. This unassisted comparison had a material bearing on the verdict. On this basis alone, also, I would set aside the verdict. [@18-19].

Bornyk is an important and helpful decision. In overturning the acquittal, the Court of Appeal has recognized some very critical and principled points about the role of a trial judge and the treatment of expert evidence.

First, the Court recognized that the trial judge must limit his consideration to the evidence before the court.

Second, the Court recognized that “expert” consideration of the evidence must be done by the expert.

While it is fair to say that expert opinion evidence should be analyzed critically and may need to be considered carefully, Bornyk illustrates that such analysis and consideration must be done properly; triers of fact are limited to a consideration of the evidence available to them.