The judgment of Mr Justice Jay in ARB v IVF Hammersmith Ltd [2017] EWHC 2438 (QB) is one that has already made headlines.  There is much of interest. However, that  part of the judgment that deals with the analysis of expert evidence is unlikely to make the headlines. It contains much of interest in relation to the admissibility of expert evidence. It also highlights the difficulty of a party seeking to challenge expert evidence when that party has not called expert evidence of its own.



The claimant (“ARB”) brought an action against a clinic alleging that they allowed his former partner (“R”) to use frozen embryos without his consent.  A child had subsequently been born and the claimant sought damages on this basis.  One key factual issue was whether R had forged ARB’s signature on the consent form.  The defendant was the clinic. The clinic, in turn, had brought in R as a third party.

The claimant (but apparently not the defendant) adduced expert evidence. The claimant’s expert was of the view that the signature had been forged.

(The outcome of the case was that the claimant’s evidence was accepted in all material respects. However the defendant clinic was not in breach of duty. Further this case came within the principle that a birth of a healthy child did not give rise to a claim for damages.)



The Evidence of Dr Audrey Giles, BSc PhD
    1. Dr Giles is a member of the Chartered Society of Forensic Sciences. Formerly, she was head of the Questioned Documents Section of the Metropolitan Police Forensic Science Laboratory. She has nearly 40 years’ experience in this domain. It is clear from her report that Dr Giles is eminent in her field, both in the UK and internationally.
    2. Dr Giles examined the “questioned signature” on the Consent to Thaw form against originals of ARB’s signature and various copies of it. In her opinion, the questioned signature demonstrates a similarity in overall style when compared with ARB’s undisputed signatures, but there were clear differences in terms of the fluency and the number of pen lifts. Dr Giles has explained to me in some detail the respects in which this is so, particularly in relation to the “R” and the “B”. Further, the final long upward stroke of the questioned signature is not present in the comparison signatures.
    3. In short:
“The questioned signature on the Consent to Thaw form lacks fluency. This signature bears a superficial resemblance to the undisputed signatures of the male partner provided to me but differs from them in detail.
Examination of the questioned signature under specialised lighting conditions shows the presence of pencil lines closely associated with the black ballpoint pen ink lines of the signature. The nature of the pencil lines is entirely consistent with their having been used as guide lines in the process of tracing. Signatures simulated by means of tracing are commonly associated with poor fluency and they contain structures which, whilst appearing to be pictorially similar to those in genuine signatures, are drawn incorrectly compared to those in the genuine signatures.
The features of the questioned signature that I have observed amount to very strong positive evidence to support the view that the questioned signature on the Consent dated 20th October 2010 is not a genuine signature of the male partner but is an attempt to simulate the signature by means of tracing.
I cannot exclude the possibility that the questioned signature is a genuine signature of the male partner made under highly unusual circumstances. However, I consider this possibility to be extremely remote.”
  1. In her oral evidence (cross-examination, not further evidence in chief) Dr Giles expanded on the technical aspects of her testimony. Under the Video Spectral Comparator, 1,000 nanometres of infra-red light shows the presence of carbon, appearing as black in colour to the human eye. Taken in isolation, Dr Giles agreed that this did not necessarily show the presence of pencil markings because it could be ink. However, she added that whereas black ink fluoresces, pencil does not – and there was no fluorescence she could see. Under the stereomicroscope, Dr Giles could see the characteristics of graphite deposits. In her experience, these were clearly the appearance of pencil markings.
  2. Mr McDonald valiantly cross-examined Dr Giles to the best of his ability, but the upshot was that her evidence was, if anything, fortified by that process. Dr Giles did not accept that her methodology was “unscientific” or “subjective”, although she did agree that this discipline is qualitative, not quantitative. She agreed that a person’s signature will show variability from time to time, and in different environments. She added that ARB’s is a sophisticated signature which is difficult to simulate. Dr Giles could see some evidence of smudging of the pen signature, which was consistent with someone attempting to rub out the pencil tracings. It is not possible, she said, for ARB’s pen to have malfunctioned part way through writing his signature, not least because her specialist examination had revealed the existence of a complete, now effaced, signature under the ink signature which is present on the document. Further, someone writing over a traced signature would find it difficult to attain the characteristic fluency of the true signatory.”



“My Assessment of Dr Giles
    1. In relation to the allegation of forgery, I agree with Mr McDonald that the burden of proof lies on ARB and I must apply the civil standard of proof (see B (Children) (Care Proceedings: Standard of Proof) (CAFCASS Intervening) [2009] AC 11 and In re S-B (Children) (Care Proceedings: Standard of Proof) [2010] 1 AC 678) to the entirety of the evidence before me, including the inferences to be drawn from that evidence and the inherent probabilities. I accept Mr McDonald’s submission that were I to decide that ARB’s evidence were not credible, that would require me to scrutinise Dr Giles’ evidence with even greater care. This, in other words, is a composite assessment which does not depend on my chosen point of embarkation.
    2. However, for the purposes of setting out my conclusions in this Judgment I do have to start somewhere, and it is convenient to begin with Dr Giles’ evidence. She was an excellent witness who gave her evidence in a careful and measured fashion notwithstanding the tone of parts of Mr McDonald’s cross-examination. I cannot fairly say whether it was Mr McDonald’s intention to ruffle the witness; if it was, he did not succeed.
    3. Mr McDonald submitted that there are a number of reasons for caution before accepting Dr Giles’ evidence, not least because (1) the veracity of handwriting analysis and the comparison of signatures has been doubted across the world, and (2) the parts of Dr Giles’ report relevant to the tracing aspect and alleged presence of pencil on the Consent to Thaw form lack detail.
    4. Mr McDonald has submitted three reports which he observed may be of interest to the Court when evaluating Dr Giles’ evidence. The UK Association of Forensic Science Providers sets out a recommended approach to the evaluation of forensic evidence. The National Academy of Sciences in the US has reviewed and highlighted some of the criticisms of the forensic sciences. Finally, an Australian study published in 2002 has examined the error rates in forensic handwriting analysis.
    5. Further, Mr McDonald characterised Dr Giles’ evidence as “subjective” (in relation to her comparative analysis of the signature on the Consent to Thaw form with ARB’s verified signatures elsewhere) and as “an assertion” (in relation to the pencil marks). In particular, it is complained that Dr Giles gave additional evidence about the pencil markings which was not in her report. Thus:
“It is not hard to see that the potential identification of something which on the face of it appears to be a pencil trace of a signature, provided an easy route to the conclusion that the signature was a forgery. This becomes even more likely when considering that Dr Giles was instructed to consider whether there was a forgery or not; creating a bias from the outset of the exercise.”
  1. I simply cannot accept these submissions. Dr Giles’ evidence is admissible as expert opinion evidence on well-established principles: see, for example, the decision of the Supreme Court in Kennedy v Cordia (Services) LLP [2016] 1 WLR 597 (in particular, paragraphs 41-44). It is the duty of an expert to be objective, to approach the given exercise without preconception and solely with regards to the available evidence. In that regard, an expert is little different from a judge. Mr McDonald’s cross-examination of Dr Giles failed to demonstrate that she violated this fundamental aspect of her duty. I agree with Mr Mylonas that it is unnecessary to characterise handwriting expertise as “scientific” in order to render Dr Giles’ evidence admissible. An expert witness may give evidence of her own observations, as well as opinion evidence based on her knowledge and experience of a subject-matter. However, the more that it is demonstrable that an expert has applied scientific methods to her task, the greater the weight that should be accorded to her product.
  2. The approach to expert evidence in other common law jurisdictions is of some interest but I am not writing a treatise in comparative law. For example, the decision of the US Supreme Court in Daubert v Merrell Dow Pharmaceuticals [1993] 509 US 579, 589 is based on social and policy considerations intimately tied up with jury trials in civil cases.
  3. In my judgment, it is clear that Dr Giles has applied scientific methods to the task of ascertaining what, if anything, was under the biro signature on the Consent to Thaw form. Mr McDonald complains that she gave greater detail about this in the witness box than she had done in her report. That is correct, but experts often do that when pressed in cross-examination. Mr McDonald might have had a valid complaint had Mr Hyam sought to elicit supplementary evidence by way of examination in chief, but he did not. The upshot is of Mr McDonald’s own making. Further, it seems to me that Dr Giles was quite entitled as an expert to unify her scientific observations with her considerable experience of similar findings.
  4. In his closing submissions but not when cross-examining Dr Giles, Mr McDonald sought to persuade me that there was nothing under the biro signature at all: in other words, what the infra-red light reveals as a palimpsest is non-existent. I do not consider that it is open to Mr McDonald to advance that submission at this stage. Its implications are that what Dr Giles has told the Court is something under the visible signature in biro is in fact that very signature illuminated in a different way. If that were really Dr Giles’ evidence, she would be seriously misleading the Court. In fact, it is not her evidence because she told me in terms that ink and pencil fluoresce differently, and she also referred me to what was visible on the stereomicroscope. Further, if one carefully examines the middle signature (item (b)) in her report, showing both the pencil and the ink signatures, it is apparent that they do not precisely overlap. It is also clear that there is no separate colouration under R’s signature. Finally, Dr Giles was not asked to explain the difference between fluorescence at the green and blue ends of the light spectrum.
  5. Dr Giles’ comparative evaluation of the signature on the Consent to Thaw form with ARB’s known “true” signatures is less scientific, and more opinion-based and experiential. Dr Giles has made a series of qualitative assessments, some of which are not particularly concrete: for example, the degree of fluency. This is not to undermine her judgments; it is merely to point out that they amount to matters of opinion which cannot be quantified. Caution needs to be exercised by me in the weight to be accorded to such matters. However, Dr Giles has also identified clear differences between the disputed signature and ARB’s “true” signatures about which in my view I need be less cautious, albeit always appropriately careful and analytical.
  6. In my judgment, Mr McDonald signally failed to undermine Dr Giles’ opinions in these respects. All that he was able to do was to point out a number of general methodological difficulties. I do not accept that “the veracity of handwriting analysis and the comparison of signatures has been doubted across the world”. That is to put the point far too high, and cannot survive Dr Giles’ direct rebuttal of it. Some weight should be given to the experience of other common law jurisdictions, but it is difficult to specify the exact evidence-base for the concerns expressed. It may well be that other jurisdictions do not routinely receive expert evidence in this field from witnesses of the calibre of Dr Giles.
  7. It follows that I must reject all of Mr McDonald’s criticisms of Dr Giles’ evidence. I have indicated the respects in which I approach her evidence with a degree of caution. Overall, however, I felt when listening to her evidence that I was in safe hands.
  8. R cannot get away from the fact that underneath the biro signature on the Consent to Thaw form is another complete signature, visible not to the naked eye but only with specialist equipment. Dr Giles has clearly explained the basis for her conclusion that this complete signature was written in pencil. We know, because it is no longer there, that it has been removed in some way. The irresistible inference is that someone has rubbed out this signature, leaving just the biro signature. For that someone else to have been ARB, he would have had to sign first in pencil (or some carbon-based product with similar properties), sign in biro above it, and then rub out the pencil. It was not put to ARB that he did that, and no remotely sensible or intelligible reason for him so doing has been advanced. It matters not for these purposes which copy of the Consent to Thaw form ARB was asked to sign, or indeed the precise circumstances which obtained before and during this event. On the other hand, someone forging ARB’s signature had a very good reason for tracing it out in pencil before signing it in biro. This “someone” could only have been R.
  9. The evidential force of this point is fortified when consideration is given to Dr Giles’ evidence, which I accept, that there are differences between ARB’s true signature and the signature on the Consent to Thaw form; and that the latter lacks fluency.
  10. In my judgment, Dr Giles’ evidence amounts to very strong evidence that R forged ARB’s signature on or about 20th October 2010. However, that evidence does not exist in isolation. It is strengthened by other evidence in this case in respect of which I make the following findings.”