/ 10 March 2014

Pistorius trial: The trouble with expert witnesses

Oscar Pistorius.
Oscar Pistorius.

Forensic pathologist Gert Saayman is an excellent witness even measured against his expert counterparts: he clearly differentiates between observation and interpretation; he uses precise medical language but explains it where necessary; and he provides detailed caveats on what his branch of science can and can not tell a court.

His evidence in the murder trial of Oscar Pistorius on Monday was, in that, fairly unusual; South African courts typically have far worse reports on autopsies (if they are carried out at all) to deal with in murder and culpable homicide cases.

But by every other measure there was nothing unusual about what Saayman told the high court in Pretoria, nothing of a particularly graphic nature, no details so horrible that they would come as a shock to those of average impressionability. Had it not been for the intense media interest in the Pistorius trial – which could come with soundbites from Saayman recycled on news bulletins throughout the day – there would have been no plausible reason to deny the public direct access to Saayman's testimony.

Yet, as Judge Thokozile Masipa said on Monday, the trial of Pistorius is not the usual state of affairs, and so real-time reporting was banned, as was extensive direct quotes from Saayman. Or so we assume.

Masipa specifically banned live-blogging and live-tweeting, but specifically allowed paraphrased and summarised reporting on Saayman after the fact. Where the middle ground lies, that is simply unclear.

In the final analysis, though, the evidence Saayman presented is almost sure to be widely misunderstood. It may have been less misunderstood had it been reported directly, and it may have been less misunderstood had those most intensely interested in the trial been allowed to see him deliver it. Yet by its very nature, the kind of testimony he provided is packed with caveats and uncertainty, of the kind that easily falls by the wayside, and also needs to be broadly interpreted within a trial rather than in isolation.

For instance, Saayman said Reeva Steenkamp probably last ate less than two hours before she died. That is of enormous importance considering that Pistorius claims the couple were already asleep by that time. Or perhaps not.

What Saayman actually said was complex. Asked what the state of Steenkamp's stomach indicated in terms of time of death, he first delivered what amounted to a mini lecture about the imprecision of any measure he could provide. He noted that this problem had been closely studied because it is both important in investigating murders and particularly fraught with difficulty. He described the imprecision of the science involved. He cited the possibility of very large meals, or very small meals, affecting the time taken to digest food. He cautioned that different types of food are digested at different speeds, that different people digest food at different speeds, and that the same person could digest food at different rates from one day to the next.

That done, and with the weight of his considerable expertise behind it, he finally allowed that Steenkamp, by his best estimate, died within two hours of last eating anything.

How will the court interpret that? If history is any guide, it will not take Saayman's evidence as either bolstering or undermining Pistorius's version of events in itself. Instead it will consider the information for what it is: another piece in a complex puzzle to be considered alongside all the other pieces in trying to determine a sequence of events.

Courts, at least some courts, are good at that. The public as a whole is considerably less so, whether watching live or reading after-the-fact paraphrased summaries.

Another case in point is the evidence of injury left on Steenkamp's body. The presence of some types of injuries, mostly to the face, often – but not always – indicates domestic abuse. As a bonus complication, the absence of such injuries do not necessarily mean the absence of abuse; showing the absence of a type of violence that is at heart psychological and secret is notoriously difficult.

Saayman, in his first round in the witness box and before cross-examination, noted none of the typical domestic-violence markers. He did, however, describe small abrasions, and the possible presence of damage inflicted through blunt trauma.

So does this mean there was not a physical altercation between Pistorius and Steenkamp on the night? Not quite. At best, the evidence presented by Saayman to date indicates that there was an altercation of the type that is easily seen, or proven. In the totality of evidence that may be useful – or it may be entirely meaningless.

In a third example, Saayman explained that the bullets that killed Steenkamp was probably (again, caveats apply) of a type designed to mushroom on impact, and so maximise the damage done to tissue.

Is that clear evidence that Pistorius intended to kill when he fired the fatal shots? Not in isolation. Young men choose ammunition based on peer recommendations, or load magazines to show off, or gun enthusiasts mistakingly believe mushrooming ammunition to be less fatal than traditional kinds. There are many reasons to argue that a choice of ammunition is meaningless in measuring intention – perhaps as many arguments as there are to the contrary.