Forensic science is a discipline that offers the curious researcher a pretty much unending supply of interesting questions that need to be explored and answered. It provides the opportunity to deal with real world challenges, to interact with a whole range of different kinds of institutions and professionals, and to actually have an impact. But Forensic science research is also often a confusing space for the very same reasons. The recent findings published by the House of Lords Science and Technology Committee found that research, as with every other part of the forensic science system, is in trouble and in need of reform (see my previous blog). So what is the issue now and where could we go?
Where are we now?
I have come to recognise that the terms ‘research’ and ‘R&D’ means very different things to different parts of the forensic science system. For example, in some areas, R&D is the term for developing tools to meet operational needs (such as a ‘lap on a chip’), whereas in others research has a much broader remit and intended outcomes (such as developing underlying theory to understand how to identify, collect or interpret science evidence and lay a foundation to inform and develop future solutions to operational needs). Both of these approaches (and all the others in between) are absolutely necessary and valuable for innovation and development of forensic science. However, these different understandings and aims for ‘research’, and the way value is both considered and assigned to the outputs of ‘research’, has meant that like with all other parts of forensic science, it has suffered from being piecemeal and lacking cohesion, overarching strategy and resourcing.
When it comes to research in forensic science there are three main areas of focus:
The scientific validity of methods used to identify the source of a material or mark, and the related challenges of addressing mixed provenance samples,
The need to understand how and when materials are generated, transferred and persist (questions of activity),
The importance of appreciating the role of human judgement and decision making at every stage from crime scene to court, identifying factors that can affect judgements (including issues around bias, use of stats and the presentation of findings to investigators or to the court), and the way of conveying the significance of findings in a case context.
Forensic science has come under strong scrutiny in the last ten years, but one area that has been particularly identified as problematic has been the underlying scientific validity of many of the techniques and approaches that are routinely used in forensic reconstructions. This is an area that has attracted much debate and argument, particularly around how we understand the term ‘unique’ and the implications that has for identifying ‘matches’, how we incorporate the scientific method into comparison approaches (for example fingermark and footwear pattern comparisons), and how we can consider ‘error’.
A lot of this scrutiny has been valuable but it is also important to consider that the paradigm for replicability and accuracy in one domain may not be straightforwardly applicable to another. It is possible to get an error rate for the analysis of a substance by mass spectrometry, because it is possible to run standards, and to run the same sample multiple times and compare the outputs. However, when it comes to a comparison of patterns, this is a different type of task and it is not possible to offer (in the same way) an ‘error rate’ of the analysis because the nature of analysis in a pattern recognition approach is different to the analysis of the components in a substance.
What has not been scrutinised to nearly the same degree, is the gap that exists in forensic science when it comes to understanding how and when traces transfer and persist. This is absolutely critical to being able to understand what it means when we find a suspect’s DNA on a gun or gunshot residue on their clothing (for more on this see my TED talk). Increasingly we are seeing that where miscarriages of justice have been identified, there was a reliance on scientific evidence to offer insights into how and when a trace was transferred, but without the necessary evidence base existing to underpin those kinds of assertions.
Then there is the absolutely critical role of humans in reconstructing forensic science events which is clear at every stage from crime scene to court. In forensic science there have been concerns identified about how decision making can be subject to biases (as it is in all decision making in every walk of life), and the (mis)use of statistics to convey significance. These are really important challenges that need addressing, but the answer is not about removing the human from this process. I’m really excited about technology and the new capabilities that are emerging, but I don’t think we’re going have a future without any human involvement when it comes to forensic science any time soon – human brains are able to do incredible things that we can’t replicate just yet. But bringing the two together does offer a huge amount of potential in terms of increasing the transparency in how decisions are made, sifting through large amounts of data to find the important material, and spotting the anomalies that are pertinent to a specific case.
And the future?
I’m a great believer in the mantra ‘step one, show up’. The House of Lords inquiry report offers a coherent overview of the current situation, and the gaps that need filling. It also highlights for the first time the different aspects of forensic science research that are needed going forward to tackle those gaps in a holistic way, as well as considering the importance of having an infrastructure to enable that in terms of a National Institute (see here for more on this).
But I’m also a great believer in the idea that innovation first needs a well-articulated problem. In forensic science we are increasingly able to articulate the ‘problems’, so we have a fertile ground for creative thinking and developing the solutions. To prolong the metaphor – what we need now is strategic sowing and watering, and then the potential for an incredible, diverse and cross-pollinating meadow will be realised for the benefit of the whole forensic science ecosystem. The right kind of National Institute has the potential to ensure that forensic science flourishes.
