A closer look at George Hutchinson

[Fisherman]

Reading expert opinions can sometimes require it's own expertise! Experts phrase things in ways that can sometimes sound very different to the untrained ear. For example, even if their opinion is on the "you may draw this conclusion safely ...", they will phrase things in terms of "it is highly probable" or "highly improbable" ... leaving the untrained ear to presume the expert is saying there is reasonable room for the alternative. Generally, that's not going to be the case and the expert is simply acknowledging that given the nature of any data analysis, nothing is 100%, and so we phrase things that way because we are trained to phrase things that way.

When an expert spends a lot of time pointing out that the data they have been given to work with is not sufficient (i.e. I would require original material, not images, etc), then they are indicating that what follows is not their expert opinion on the matter, rather, they are simply being polite and trying to suggest what such an analysis might show, if they indeed had the proper, and sufficient amount, of material to do an analysis upon.

Finally, it is not uncommon for people to think that experts will always offer an opinion of "yay or nay", but there is a third, often more common opinion that emerges, which is in between - that's the "equivocal" outcome, meaning it is unclear if the answer should be yay or if it should be nay, as the evidence could fit either of those. It's not sufficient to determine. So where he is talking about "can not be ruled out", he's indicating that such a phrase is an expression from the "equivocal outcome"; it can't be ruled out does not mean it has been ruled in. We are in that grey area of either is possible.

Given he clearly indicates that the material he working from is not such that he could (or will) form an expert opinion, and that he goes on to try and explain how saying "can not be ruled out" simply means that "there are obvious likenesses in certain respects", which is a far cry from saying "the similarities are such that we can be confident the same hand has a high probability of the being source of both signatures" (the latter being a phrase one might seen if an expert was of the opinion there was a match, the former being an indication of non-committal opinion).

I have no experience with handwriting analysis myself, but have lots of experience with expert opinions and how we phrase things (I'm an expert in some areas of research, most of which bear little relevance to criminology, but expert language is a function of how we're trained to talk about data, and analysis, and statements based upon the probabilities derived from statistical analysis and so forth, so I feel somewhat confident in sharing my thoughts here). Researchers are a funny lot, and we end up with idiosyncratic ways of expressing things that can easily be misinterpreted by those not dealing with data and statistical analyses on a day-to-day basis.

And also, that training, and how experts "speak", has changed since the 1800s of course. Experts were expected to speak in more definite language, but over time, things have shifted to use language that more accurately reflects the fact that even if the analysis leads you to be 99% sure, you cannot say definitely, because, 1% error means you're wrong 1% of the time after all, while definitely implies there's 0% room for error - that you cannot be wrong. In such a case, the expert would resort to describing things something like "In all reasonable probability the two items were written by the same hand", but if questioned by a defence lawyer, they would have to admit that "Yes, there is a 1% chance that the signatures were written by two different people", and it would be the jury that would have to decide if 1% was reasonable doubt.

- Jeff

If, as you say, a jury would have to admit that there must be a one per cent chance that two signatures deemed similar were by different hands, then the matter of how both men in our case also happened to use the exact same name when signing would put that self same jury at rest. It would go from a nearly certain call to a very certain one.

[rjpalmer]

If, as you say, a jury would have to admit that there must be a one per cent chance that two signatures deemed similar were by different hands, then the matter of how both men in our case also happened to use the exact same name when signing would put that self same jury at rest. It would go from a nearly certain call to a very certain one.

Isn't this somewhat misstated, Christer? It's as if you are reversing the order of things for dramatic effect.

It's not like the claimant didn't already know that the witness's name in the Kelly murder was the same as his father's name.

You make it sound like the two signatures were matched and then...lo and behold...it was noticed that the two men had the same name, too, making it a 'very certain one.'

In reality, the fact that Reggie's father's name was the same as the Kelly witness is what allowed him to claim they were the same person. That element was a given--a necessary ingredient in the claim.

It was then noticed the signatures were indeed similar.

Now, granted--if the two signatures looked nothing at all alike, Reggie's claim could have been instantly dismissed as hot air.

That they looked similar (but also have some dissimilarities, as your expert notes) makes the claim interesting.

No disagreement.

But the way you state it in reverse order is overstating the 'certainty' of it.

You're creating an extra coincidence that isn't there. If his father's name had been Joe Bloggs, the comparison would have never been attempted.

[Fisherman]

Isn't this somewhat misstated, Christer? It's as if you are reversing the order of things for dramatic effect.

It's not like the claimant didn't already know that the witness's name in the Kelly murder was the same as his father's name.

You make it sound like the two signatures were matched and then...lo and behold...it was noticed that the two men had the same name, too, making it a 'very certain one.'

In reality, the fact that Reggie's father's name was the same as the Kelly witness is what allowed him to claim they were the same person. That element was a given--a necessary ingredient in the claim.

It was then noticed the signatures were indeed similar.

Now, granted--if the two signatures looked nothing at all alike, Reggie's claim could have been instantly dismissed as hot air.

That they looked similar (but also have some dissimilarities, as your expert notes) makes the claim interesting.

No disagreement.

But the way you state it in reverse order is overstating the 'certainty' of it.

You're creating an extra coincidence that isn't there. If his father's name had been Joe Bloggs, the comparison would have never been attempted.

It is of course not about whether our not Reg Hutchinson knew that his fathers name was George Hutchinson. It is about how a document examiner says that the signatures of Topping Hutchinson and the witness of Ripper fame are similar. Once we know that, the suggestion that somebody OTHER than Topping Hutchinson was the killer requires not only that this somebody else had a signature that was similar to that of Topping, but ALSO that he by pure chance happened to have the same name or moniker.

Mentioning that a son is likely to know his fathers name is immaterial in that light.

If we turn it around, and reason that Reg Hutchinson thought that he was going to offer up his fathers name as the likely witness, after having noted that Toppys signature was remarkably similar, we still have the same enigma: Why would the two men have so similar signatures AND the same name, if it was not because they were the same man?

The dramatic effect is not something I make up - it is inherent in the matter. The exact same name name, similar signatures - but different men. That would be about as dramatic as it gets. Claiming that I am creating an extreme coincidence that is not there is not true: BOTH factors are there, the name AND the similarity in writing style.

[rjpalmer]

The dramatic effect is not something I make up - it is inherent in the matter.

If you can't appreciate the distinction, Christer, then I don't know what to tell you.

Go back and read what you wrote and reflect on how it is misstated.

If the handwriting analyst studied the handwriting of everyone alive in London in 1888--regardless of their name--and found a 'match' and only then found out that the witness's name was also George Hutchinson (and this is what you are implying in your post) then that would indeed make it quite certain.

But that's not what happened. Not in the least.

[Fisherman]

If you can't appreciate the distinction, Christer, then I don't know what to tell you.

Go back and read what you wrote and reflect on how it is misstated.

If the handwriting analyst studied the handwriting of everyone alive in London in 1888--regardless of their name--and found a 'match' and only then found out that the witness's name was also George Hutchinson (and this is what you are implying in your post) then that would indeed make it quite certain.

But that's not what happened. Not in the least.

Well, R J, if you cannot accept how there are TWO matters involved that amplify each other, then I don´t know what to tell you.

As for Frank Leander only finding out that the witness´name after he had compared the signatures, he is not in any shape or form a ripperologist. I sent the two signatures too him for comparison, and I did not tell him what they were about until I had word from him that they were similar.

The two factors taken together are enough to put it beyond any doubt that the witness was Topping Hutchinson, it is that simple.

An easy exercise:

1. Do we or do we not have two signatures, one by the witness of Ripper fame and one by Topping Hutchinson?

2. Are these signatures similar or are they not?

3. Once we know that these two signatures are similar, it applies that Reg Hutchinson many years ago, stated that the witness was his father, George William Topping Hutchinson. And since we have TWO, not one, similarities (name AND signature style), there is every reason to regard the matters as cleared up.

What is the alternative? That Reg Hutchinson heard about the witness testimony and the three signatures, and falsified his fathers signature in order to make it look alike the witnesses signature? That would be the only risk we are running - and we are not running it, because we do have Toppings signature on record, from when he was married, from when he baptized his children etcetera. And lo and behold, they ARE similar to the signature of the witness. That is one of the two parameters in place.

The second parameter is the similar names. And here it applies that we know that we have the same name on all the documents involved.

Once we know this, and if we want to claim that the man who signed the witness testimony was perhaps NOT Topping Hutchinson, we have two obstacles to overcome:

1. Both men had the same name (unless the witness lied about his real name and used George Hutchinson as a moniker only). This is where you seem to think that it carries no weight that the names were the same, but I must beg to differ.

2. Both men, carrying the same name, wrote in a similar enough style for a forensic examiner to say that he expected any forthcoming evidence to further strengthen the suggestion of a common originator.

Each of these matters may be coincidental. But both of them are unlikely in the extreme to be.

If I am making too good a case, I can only say I am sorry.

[Sam Flynn]

3. Once we know that these two signatures are similar, it applies that Reg Hutchinson many years ago, stated that the witness was his father, George William Topping Hutchinson. And since we have TWO, not one, similarities (name AND signature style).

More than two similarities, in fact; Toppy lived in south-east London and had family connections to Essex, as did witness George Hutchinson. There's also the family tradition identifying Toppy as said witness... which, granted, is hardly concrete proof, but I'm not aware of anyone else claiming their ancestor was "our" Hutch.

[caz]

Hi Christer,

I think I would want to compare the signatures of a hundred different men, all named something like Edward Smith, who were educated in Victorian times, to establish how differently or how similarly each formed the individual letters of their name.

That way, one might get a better feel for what percentage would have looked very similar by chance alone in that era, due to the way people were taught to write, and what percentage would look too different to have been by the same Edward Smith.

I would expect to see more similarities than differences, but without doing the experiment, who knows?

That said, Sam makes a good point about the circumstantial similarities in Toppy's case.

Love,

Caz
X

[Fisherman]

Hi Christer,

I think I would want to compare the signatures of a hundred different men, all named something like Edward Smith, who were educated in Victorian times, to establish how differently or how similarly each formed the individual letters of their name.

That way, one might get a better feel for what percentage would have looked very similar by chance alone in that era, due to the way people were taught to write, and what percentage would look too different to have been by the same Edward Smith.

I would expect to see more similarities than differences, but without doing the experiment, who knows?

That said, Sam makes a good point about the circumstantial similarities in Toppy's case.

Love,

Caz
X

I hear a lot about how the Victorians were all taught to write in the same way. In order to check that out, I have taken a look in contemporary hotel ledgers. The fact of the matter is that they were just as disparate as we are today.

Take a look yourself, an d you will see what I mean.

[Fisherman]

More than two similarities, in fact; Toppy lived in south-east London and had family connections to Essex, as did witness George Hutchinson. There's also the family tradition identifying Toppy as said witness... which, granted, is hardly concrete proof, but I'm not aware of anyone else claiming their ancestor was "our" Hutch.

True, and good points. On the whole, I don´t think more is needed to rule out any other signer than Toppy for the witness signature, but there is nothing wrong with over-over-overwhelming evidence!

[rjpalmer]

More than two similarities, in fact; Toppy lived in south-east London and had family connections to Essex, as did witness George Hutchinson. There's also the family tradition identifying Toppy as said witness... which, granted, is hardly concrete proof, but I'm not aware of anyone else claiming their ancestor was "our" Hutch.

Yes, but using the 'same name' argument as Christer is doing is circular.

It's not, in itself, evidence; looking at it objectively, it's merely the circumstance that led to the claim being made in the first place, so it can't be cited as independent evidence.

After all, we are talking about an unidentified person in a notorious criminal case (a ballpark analogy in the United States would be the unidentified 'D.B. Cooper' of hijacking infamy) so it's hardly out of the question that someone whose ancestor had the same name would come forward and claim that their father or grandfather was the same George Hutchinson.

That the two men ended up having very similar signatures can be considered evidence. I readily acknowledge that--it's certainly a mark in Toppy's favour.

But the same name? The 'family tradition' could be poppycock.

Maybe there are more b.s. artists in the U.S. than in the UK, but there have been dozens of men who have claimed their father or uncle was D.B. Cooper. I wouldn't consider their name being Cooper as evidence.

And I say this even though I tend to think Toppy could be the right man.

At the same time, there's a lot of similarity in handwriting. I once snipped the following men out of the 1911 UK Census just for jolly, making sure they were all born 1855 +/- 10 years.

Speaking strictly as an amateur, I would exclude some of them, but I'm not sure I would exclude all of them.

[rjpalmer]

1. Both men had the same name (unless the witness lied about his real name and used George Hutchinson as a moniker only). This is where you seem to think that it carries no weight that the names were the same, but I must beg to differ.

Think of it this way, Christer.

Now and then people have come forward to claim that their ancestor was the unidentified Mary Kelly of Miller's Court.

Chris Scott once encountered one of these people and researched it, showing that their family tradition was wrong.

So, was the 'same name' a good argument or evidence?

No, because the 'same name' is what led to the claim and/or the family tradition in the first place.

If we had Mary Kelly's signature and their ancestor's signature had similarities, we could consider this evidence. The name, in itself, is not.

That's all I'm saying.

P.S.

I suppose you could argue that the ancestor's name being 'Mary Jane Kelly' would make it a more plausible claim than someone who's ancestor was Mary Jane O'Connor, but until the victim (or the witness) is positively identified, that argument might be considered less than safe.

Semantics.

[JeffHamm]

If, as you say, a jury would have to admit that there must be a one per cent chance that two signatures deemed similar were by different hands, then the matter of how both men in our case also happened to use the exact same name when signing would put that self same jury at rest. It would go from a nearly certain call to a very certain one.

Hi Fisherman,

I didn't say the jury had to "admit" that there must be a one per cent ... etc. It would be the expert who would have to admit there was a chance their opinion was incorrect, even if they deemed the probability of that was so small they were willing to state that in their opinion the two signatures were written by the same person. And in some cases, the Hitler Diaries for example, where the expert was fooled and made a very strong, positive identification, and yet, they were wrong. An expert draws a "firm" conclusion when they deem the probability of the alternative answer has become sufficiently small - but sufficiently small isn't zero, so the expert has to "admit" the chance for error is not zero. That's where you get things like DNA experts testifying to things like "the probability that this sample came from anyone else is 1 in 5 billion" - they are "admitting" there is a 1 in 5 billion chance they are wrong. Juries tend to decide that does not reach the level of reasonable doubt.

As I had said, the jury, when presented with the testimony have to then decide if the small margin of error meets the criterion of reasonable doubt (they don't have to "admit" anything, they only have to decide based upon what was presented to them). That becomes more likely should the expert's testimony on their analysis not allow for a definitive statement, as in our example being looked at here where the expert admits that images are not source material that an analysis can even be properly undertaken, hence they specifically say they are not offering an expert opinion.

There are some areas of forensic analysis that have come under scrutiny in terms of their reliability, in part due to the lack of standard methods and procedures for the analyses, and with a lack of reliable standards. Bite mark analysis, for example. The field suffers from lack of standardized protocols and analyses. I recall seeing a short documentary on that area, where they wanted to do a study to determine the reliability of experts in the field. Things like, send a bunch of bit mark evidence, and see what percentage of the experts all identified the matches as matches, the mismatches as mismatches, and identified the foils (non-bitemarks) as foils. They never got to the reliability of the match/mismatch question because the performance on simply identifying which photos actually depicted bite marks was pretty much at chance levels. This was prompted because, one of the leading experts was getting frustrated that so many experts were willing to do this sort of thing based upon photos, and he was adamant that you cannot make any judgement from a photo - you have to examine the body itself. His concern was validated from what they found.

I'm not sure how many reliability studies have been done on handwriting comparisons, or if there are standard protocols by which the comparisons are made, and standardized criterions for determining the likelihood of a match or not. If much of the analysis is individualized (each expert has their own "way", bases "these bits are similar" and "these bits differ" on solely their own opinion and not based upon standardized measurement ranges, and so forth, then I would be very concerned because under such loose conditions then it becomes very probable that if you ask a different expert you would get a different opinion.

Even in an area I've done some of my own research, spatial analysis of crime scene locations, I was concerned about the lack of standardized procedures. In the paper I published on it, one of the issues I focused on was how accurate were some of the procedures used to draw inferences, and I demonstrated that some of them are completely bogus while others were reliable. I suggested a number of steps and procedures that, if followed, would help to standardized the analysis steps to aid the investigator (or researcher) when it comes time for them to make an inference.

In the end, expert opinion is only as good as the field allows. If the field is chaotic, with no formal approaches, then it becomes increasingly unlikely that two experts will draw the same opinion because each expert does something different, bases their opinion on different criterions, and so forth. Some experts in a field may be highly reliable and accurate, others may be less so, but without studies examining such questions, it's hard to know who is which.

Caz's suggestion for a study would be a great first step. That's exactly the sort of thing that has to be done, in many fields.

- Jeff

[rjpalmer]

I'm not aware of anyone else claiming their ancestor was "our" Hutch.

Okay, fair enough.

I suppose we could consider this 'negative' evidence.

Only one person has claimed to be a descendant of George Hutchinson.

By contrast, there were two claimants swearing to be one of the surviving twins in the Tower--Perkin Warbeck and Lambert Simnel--and several claiming their ancestor was Mary Kelly.

So, I suppose it might carry some small weight that only one person has ever come forward.

[Fisherman]

Think of it this way, Christer.

Now and then people have come forward to claim that their ancestor was the unidentified Mary Kelly of Miller's Court.

Chris Scott once encountered one of these people and researched it, showing that their family tradition was wrong.

So, was the 'same name' a good argument or evidence?

No, because the 'same name' is what led to the claim and/or the family tradition in the first place.

If we had Mary Kelly's signature and their ancestor's signature had similarities, we could consider this evidence. The name, in itself, is not.

That's all I'm saying.

I know exactly what you are saying, R J.

Nevertheless, the name of the witness and the name of Toppy was the same. And regardless of how Reg Hutchinsons story is chronologically later than the signing of the witness, the fact remains that back in 1888, there were - according to some - two men who both signed themselves George Hutchinson, one on the witness testimony, and one of Topping Hutchinsons marriage certificate and census listings. This is the suggestion made by those who like the witness for the killers role. However, once we know that both the witness and Topping signed themselves in a fashion that was similar, we must accept that there is every reason to believe that they were the same man. And since we know that they also signed themselves George Hutchinson, both of them, you need not be a genius to understand that two different matters are both in sync. If the witness had signed himself Carl Calamity, it would still be likely/possible that he was Topping, because he wrote in a very similar fashion. But once he signed himself George Hutchinson, we get a watertight case.
You seem to think that much as we should accept that the similar writing style points to a possible/likely identification, we should not care about how the two also had the same name. And you present the Kelly material above.
Just as you say, two people can have the same name, and therefore, uncritically accepting that Mary Kelly 1 must be Mary Kelly 2 is plain dumb.
But if it could be shown that the two Mary Kellys had similar signatures as per a document examiner, that would clinch this case too. It is just how it works. And it has nothing at all to do with anybody exagerrating. It is instead all about how the two factors writing style AND name cannot both be in line without it being conclusive enough evidence.

Think of it THAT way, J R.

[Fisherman]

Hi Fisherman,

I didn't say the jury had to "admit" that there must be a one per cent ... etc. It would be the expert who would have to admit there was a chance their opinion was incorrect, even if they deemed the probability of that was so small they were willing to state that in their opinion the two signatures were written by the same person. And in some cases, the Hitler Diaries for example, where the expert was fooled and made a very strong, positive identification, and yet, they were wrong. An expert draws a "firm" conclusion when they deem the probability of the alternative answer has become sufficiently small - but sufficiently small isn't zero, so the expert has to "admit" the chance for error is not zero. That's where you get things like DNA experts testifying to things like "the probability that this sample came from anyone else is 1 in 5 billion" - they are "admitting" there is a 1 in 5 billion chance they are wrong. Juries tend to decide that does not reach the level of reasonable doubt.

As I had said, the jury, when presented with the testimony have to then decide if the small margin of error meets the criterion of reasonable doubt (they don't have to "admit" anything, they only have to decide based upon what was presented to them). That becomes more likely should the expert's testimony on their analysis not allow for a definitive statement, as in our example being looked at here where the expert admits that images are not source material that an analysis can even be properly undertaken, hence they specifically say they are not offering an expert opinion.

There are some areas of forensic analysis that have come under scrutiny in terms of their reliability, in part due to the lack of standard methods and procedures for the analyses, and with a lack of reliable standards. Bite mark analysis, for example. The field suffers from lack of standardized protocols and analyses. I recall seeing a short documentary on that area, where they wanted to do a study to determine the reliability of experts in the field. Things like, send a bunch of bit mark evidence, and see what percentage of the experts all identified the matches as matches, the mismatches as mismatches, and identified the foils (non-bitemarks) as foils. They never got to the reliability of the match/mismatch question because the performance on simply identifying which photos actually depicted bite marks was pretty much at chance levels. This was prompted because, one of the leading experts was getting frustrated that so many experts were willing to do this sort of thing based upon photos, and he was adamant that you cannot make any judgement from a photo - you have to examine the body itself. His concern was validated from what they found.

I'm not sure how many reliability studies have been done on handwriting comparisons, or if there are standard protocols by which the comparisons are made, and standardized criterions for determining the likelihood of a match or not. If much of the analysis is individualized (each expert has their own "way", bases "these bits are similar" and "these bits differ" on solely their own opinion and not based upon standardized measurement ranges, and so forth, then I would be very concerned because under such loose conditions then it becomes very probable that if you ask a different expert you would get a different opinion.

Even in an area I've done some of my own research, spatial analysis of crime scene locations, I was concerned about the lack of standardized procedures. In the paper I published on it, one of the issues I focused on was how accurate were some of the procedures used to draw inferences, and I demonstrated that some of them are completely bogus while others were reliable. I suggested a number of steps and procedures that, if followed, would help to standardized the analysis steps to aid the investigator (or researcher) when it comes time for them to make an inference.

In the end, expert opinion is only as good as the field allows. If the field is chaotic, with no formal approaches, then it becomes increasingly unlikely that two experts will draw the same opinion because each expert does something different, bases their opinion on different criterions, and so forth. Some experts in a field may be highly reliable and accurate, others may be less so, but without studies examining such questions, it's hard to know who is which.

Caz's suggestion for a study would be a great first step. That's exactly the sort of thing that has to be done, in many fields.

- Jeff

You are forgetting about the identical names on the witness testimony and Toppings marriage certificate, Jeff. I know that there must always be a possibility that a document examiner is wrong. But I also know that when not only the writing style but also the names are similar, that kind of proves that we are dealing with the same man.