John Richardson

[FISHY1118]

The problem is though Fishy what else could it have been? If Chapman was killed earlier then would could suggest someone moving around in the yard with a corpse there. And we can surely expect the police to have asked the residents if anyone had entered the yard between the time that Richardson had left and Davis finding the body?

We would then have to assume that the noise against the fence was in fact made by someone else who entered the yard , that might not nessasarily be the case . As george has suggested it could have been any number of things at that time of the morning, whos to say. I just think its a little unsafe to say categorically it was a person who created the noise .

[GBinOz]

Hi George,

Oh, absolutely the sample size is too small once I break it down into the <3 hours etc ranges, and I did mention that. Even the sample of 12 in the thesis is on the small size for my liking, but I tend to like larger than usual samples. But, for arguments sake, let's say that the equation is usually spot on, but of course with some error of estimate, that is normally distributed (so equally likely to be an over or under estimate). We have 12 cases, 11 of which are over estimates and 1 of which is an under estimate. If my above notion is true, then our outcome, or one more extreme, only has a 0.0024% chance of occurring. That's significant, meaning, I reject the hypothesis that the equation is equally likely to produce over and under estimates, and conclude the equation is prone to overestimations.

The thing is, the 4 cases that are the shortest interval (<3) show no indication of the error range narrowing, and if anything show a larger one. That's entirely understandable because the equation used is a linear function of time and cooling follows an exponential pattern over time, so the slope is constantly changing. Cooling is more rapid the greater the difference between the body and the surrounding temperature, so it starts off fast and slows as it approaches the environmental temperature. Using a linear function to model that is going to produce errors, and those errors will be exactly of the sort we observe with larger over estimations during the shortest times that decrease over time because the constant linear decrease of the model underestimates the initial fast rate of cooling but will eventually "catch up" with the cooling curve as the rate of cooling slows. That initial rate is even faster as the temperature drops, and as the thesis notes, the error gets larger the cooler the environment.

As I mentioned, I wouldn't put much stock in the average 5.6 hours overestimation value (or whatever it was) per se, and a larger sample would be required to get an idea of how big that overestimation tends to be. What we don't have, however, is any indication the error range gets smaller as the PMI gets into the 1-3 hour range. Also, given what we actually know about cooling, and what we know about any reading based model Dr. Phillips could have used, an increase in the overestimations at the shortest intervals makes perfect sense due to the fact a linear model is a crude approximation of cooling. It's not worthless, it provides a rough estimate, but it will be prone to errors. The thesis found using a faster rate of cooling improved the predictions, which makes sense as many of their cases have relatively short PMI, where the cooling rate is fast.

So if Dr. Phillips used the rate of cooling that Wickerman found (1 deg/hour), then he appears to have used a very low value for cooling, one that doubles the more accurate version (2 deg/hour) for short PMI. This allows us to recalculate his 2 hour PMI to an estimated PMI of 1 hour, and therefore his observations would be exactly what the witnesses imply as well.

I guess, given it can be shown that Dr. Phillips, if he even did take temperature readings and do a calculation, is likely to have overestimated the PMI (and may even have used a cooling rate that would double the PMI), and since we can extract what his temperature drop must have been (either 2 or 3 deg, pending upon which equation he may have used), which allows us to re-estimate the PMI using a more accurate value (2deg/hour), resulting in a PMI estimate that differs from the witnesses by 25 minutes or 5 minutes, respectively, which places the ToD after Richardson's visit. Does that make you reconsider Richardson's testimony? If not, I'm curious as to why not, particularly in light of the fact that we have actual data that shows the type of equation that Dr. Phillips would have available to him to use overestimates the PMI (that sample size of 12 is enough to demonstrate that claim, with statistical certainty).

- Jeff

Hi Jeff,

Equations have the property that they cannot be solved if there are too many unknowns. In the case of the C5 murders, were are not privy to any of the input data for the equations, and have a range of variables that may have been used. You have used Phillip's resulting PMI to work backwards and assess the input data that he would have used in the various methods of assessment that he may have adopted.....I think, have I got this wrong? Having arrived at a PMI there is a margin of error which has been shown by your examples to not conform to a bell curve, tending more towards over than under estimation. I may be misunderstanding, but I'm thinking that in a large sample the under estimate side would still be populated but to a lesser extent than the over estimate side.

What I was asking in the second part of my previous post was, using your models can a table be produced for the 8 (or 9) doctor's estimates showing the PMI's, or ToD's of the C5 murders together with their assessed error margins?

[JeffHamm]

As the'Models'used give a plus or minus error,it is equally fair to state that Phillips might have underestimated.If the error was as much as one hour,a figure that has been mentioned,then it pushes the time of death well back into the early hours.
Like I mentioned before,the elements of a theory show a possible solution.Those elements have to be proven.to give a correct answer.
One can manipulate Phillips,Long's,and Cadoche's testimony,to show a possibility agreement,but only by refusing to accept there are alternatives.

Hi harry,

For short intervals it appears the models are not equally likely to over and underestimate the true PMI but are far more prone to overestimation. Moreover, I've never presented this as "proof" the murder had to be at 5:25, and in fact I'm pretty sure I've said specifically is that given the information we have that a murder at 5:25 is the most supported hypothesis (note, hypothesis means it is still not considered a proven fact), and that other hypotheses are, at the moment, a distant second. Distant because in order to support them a lot of the information we have has to be assumed to be completely wrong, not just wrong in some details but entirely dismissed as erroneous. There is no proof that whole sale dismissal is justified, and therefore the alternative explanations require a greater culling of the information we have by assumption only. In contrast, a 5:25 murder really only requires one detail of testimony as given to be viewed as erroneous, and that is the time of Long's sighting (rather than being at 5:30 it gets presumed to be at 5:15, requiring only her to misremember which chimes had sounded when she saw the couple outside of Hanbury Street). All other "conflicts" in the multiple statements by Richardson and so forth look to be the sort of variation that emerges when comparing statements given over multiple occasions (not to mention that some of those statements may very well reflect wording of a reporter rather than the wording of Richardson himself).

In short, unless the case actually gets solved, then of course there are alternatives. But from our vantage point, over 130 years after the fact, all we can do is weigh up how the different explanations fit with the information we have, and we attempt to rank them in an order of "goodness of fit", and given how little has to be "blurred" if the murder occurred at 5:25 compared with a pre-Richardson visit murder, I simply cannot understand how anyone can view the latter as a better fit to the information we have available. I do understand the reminders that witnesses can be shockingly wrong, and therefore the alternatives cannot be dismissed entirely (and I do not dismiss them entirely; if I did I would say I've proved the murder was at 5:25, and I've made no such claim, and I re-emphasize that here).

We are, as always in this case, left with multiple possibilities, but not all possibilities can be viewed as equally probable ones. At the moment, the weight of probability lies most heavily with a murder at 5:25ish. Proof is when that probability reaches 100%, and that we do not have. All it would take is some new source of information that goes against 5:25 hypothesis (not a new "what if x was wrong", but new evidence of some sort) for us to readjust our evaluations.

As I say, the estimated PMI is considered "not in conflict" if the witness time falls within it, and it does, therefore one cannot say there is evidence of a conflict. It doesn't prove the witness time is correct, it just removes the argument that the expert opinion conflicts with the witnesses. It doesn't.

- Jeff

[GBinOz]

For me, it comes down to suspension of disbelief.

In Scenario A: we need to suspend disbelief that John Richardson overlooked a fresh corpse a couple of feet away, and Albert Cadosch heard a separate incident next door.

In Scenario B: we need to suspend disbelief that Dr Phillips' rough estimate using inexact science was off the mark.

However, by giving Richardson and Cadosch the benefit of the doubt, the corollary is that Annie and the killer were both taking a risk at that hour. Was the killer brazen enough to carry out these mutilations in broad daylight, possibly bloodstained, and set off into the madding crowd?

Hi Harry D,

A good post. Modern opinion says that in 1888 doctor's PMI's were rough estimates using inexact science. Modern opinion is that over 50% of witness testimony is unreliable due to shifting memory patterns which are evidence here by the change of the memories of the witness over time.

Your second paragraph is interesting. Was the killer brazen enough to carry out these mutilations in broad daylight, possibly bloodstained, and set off into the madding crowd? Rob Hills would say yes for his suspect, that he slipped into the cat meat shop at the front owned by his mother and cleaned up there. But continuing on your theme, how many of the outdoor canonical murders were committed in broad daylight? How many were committed were there were potential audiences in buildings surrounding the site at a time when the occupants were up and about getting ready for their day? How many were there where Jack allowed a potential witness within a few feet of him without fleeing, or allowed the same witness to return to scene with Jack just continuing his mutilations? In how many was the victim heard to cry out? The answer is in all cases, just this one. It just doesn't fit Jack's M.O.

Cheers, George

[DJA]

Nah.
Was not broad daylight.
Look at Stride.

[Fleetwood Mac]

Are you unable of comprehending the simplest of things? Show me where he ‘changed’ anything. And I don’t mean in your imagination.

Write on this.

Dear Sherlock,

It is there in the inquest testimony and has been put before you in the previous two or three pages in this thread.

Add in Long's 5.30 sighting. They have to enter 29 Hanbury Street, make their way to the yard (unseen), murder and mutilations, the murderer has to escape (unseen); all in half an hour. People in that area came alive again around 5am (work, pubs and so on). It's a stretch. Furthermore, both Long and Cadosch claimed to know the time, but they contradict one another.

It's weak, Sherlock. The case for the witnesses is weak. Only mental acrobatics, invention and manipulation of the witness statements can save the case.

Dr Phillips on the other hand, leaves us the valuable observations: 'commencing of the limbs' and 'little food in the stomach'. It doesn't take any invention nor manipulation of his statement to agree with his TOD, all that is needed is a conclusion from his observations, i.e. follow the science.

Were this a boxing context, the referee would step in, stop it and declare Dr Phillips the victor by TKO.

[Fleetwood Mac]

but added the caveat that the conditions could have meant less than 2 hours.

Which of course Dr Phillips didn't. More manipulation of the witness statements.

'At least two hours' is unequivocal.

I fear at this point, with the 'witnesses' case crumbling, you are falling back on old arguments discussed approx. 10 pages back.

We have moved on from that, we're now discussing the witness statements. We've looked at Dr Phillips, we had a good look at Dr Phillips, now it is the witnesses turn.

With the slightest bit of challenge to the witness statements, you're now deflecting, obfuscating and in Trevor's words (flannelling). This only serves to highlight just how weak the case is for the witnesses.

It is weak, Sherlock. So weak in fact, at this point the whole case for the witnesses is around 4 or 5 taps on a laptop keyboard from being utterly demolished.

Now, Long, Richardson, Cadosch:

Can you provide an argument for them that doesn't involve invention?

[Fleetwood Mac]

Too complicated for you is it.

Witness makes a trivial error - bin him!
Witness doesn’t mention something until further questioning - bin him!
Witness disagrees with a Doctors guess - bin him!
Witness disagrees with Trevor set-in-stone opinion - bin him!

Thankfully most posters don’t exist in a black and white world.

I think 'most posters' want to hear your case for the witnesses.

In points one and two above, you don't make much of a case.

1) In your own words, there are 'trivial errors'. 'Seems you are being selective in terms of what is important and what is not in the witness statements. It is the case of course that upon identifying 'errors', the witness statements by extension are compromised.

2) It's not necessarily that he didn't mention something until further questioning, the point is that you have to believe Richardson didn't think to say: "there's not much point in getting that knife, I borrowed one from the market". It defies reason.

And, what about Long and Cadosch. They both claimed to know the time. Long heard the Brewer's Clock strike half 5. What are we to do? Invent a scenario? Claim they made 'trivial errors'? In your own words they have form for 'trivial errors' and so we must take this knowledge into the reading of other parts of their statements and analyse them with a suspicion that those other parts may contain 'trivial errors' also.

It's weak, it's crumbling and at this juncture it needs putting out of its misery, Sherlock.

[Herlock Sholmes]

As the'Models'used give a plus or minus error,it is equally fair to state that Phillips might have underestimated.If the error was as much as one hour,a figure that has been mentioned,then it pushes the time of death well back into the early hours.
Like I mentioned before,the elements of a theory show a possible solution.Those elements have to be proven.to give a correct answer.
One can manipulate Phillips,Long's,and Cadoche's testimony,to show a possibility agreement,but only by refusing to accept there are alternatives.

No one has ever said that Phillips couldn’t possibly have been right Harry. The intransigence has come from posters like yourself who have continually tried to claim that Phillips must have Ben correct just because he was a competent Doctor.

[Herlock Sholmes]

[QUOTE=FISHY1118;n792150]

.





A further consultation of the detectives engaged in the case was held this morning, and an officer again visited the back-yard of No. 29, Hanbury-street, and made a careful inspection of the palings leading from that house to No. 27, where resides the young man Cadosh, who stated at the inquest that he heard sounds proceed from the spot where the body lay at a quarter-past five on the morning of the murder. An examination of the fence shows that immediately over the place in the yard there is an aperture in the palings by which the dead body could have been plainly visible, while anyone moving in the yard might easily have been seen.14 Echo Sept 20th 1888.[/QUOTE]

Just for those who might be interested, here is a ''3rd'' contemporary sketch of the back yard of 29 handbury st,

My concern with Cadosch is at 5.30am in what must have been now daylight how is it possible he didnt see the killer and Chapman through those fence pickets? . As is also the suggestion of the above newspaper article

Because the sketch is clearly inaccurate Fishy. The quote that you posted mentions ‘an aperture.’ One gap. If Cadosch hadn’t looked directly or closely at it he wouldn’t have seen through it. He would have passed it in a fraction of a second of course. If he’d seen something he’d have said that he’d seen something. If it was obvious that there couldn’t possibly have been someone in that yard without him clearly seeing them through the fence then the police would have seen this and dismissed his evidence straight away. They didn’t, so clearly it wasn’t obvious.

[Herlock Sholmes]

Hi Herlock,

I'm sure that you would have seen a nineteenth century dessert knife. I inherited a set of cutlery from my grandparents, bone handled, made in England in the early 1900's. The dessert knives have rounded points and blunt (Richardson's was also rusty and had the handle missing). They were designed to cut sponge cake or custard tarts. No one in their right mind would imagine that they could cut a tough steak let alone boot leather. He could not have cut the leather with that knife, but he said he did. Human memory can be adjusted without the custodian of that memory being aware of that adjustment. Or, he may have been lying.

Cheers, George

Hello George,

The problem is that we can’t know exactly how dull the knife was. You can still cut with a knife that isn’t particularly sharp and we have to remember that he hadn’t necessarily planned to use that knife. I’d say that it was likely that after he’d left his house that morning and started walking he found that his boot was still uncomfortable and so he decided to try and repair it at Hanbury Street where he could sit down. He had a knife on him so he would use that and he initially felt, or just hoped, that it would have been up to the job. He managed to cut some piece off but not enough so he completed the repair at the market with a sharper (and possibly longer) knife.

I realise that this is a repetition George but Richardson had no reason to lie about this and not only did he not benefit from this he actually drew more attention to himself. I’m not disputing your penultimate sentence but I really see no evidence to doubt him. I believe that Richardson was totally honest on this point. And that’s not naïveté or bias. On the opposite side to your point I think it’s also worth saying that we can be guilty of being overly cynical. Of assuming the worst in people. Of seen the sinister in everything. I can produce no statistical evidence for this George but I’d say that the vast majority of witnesses are honest (possibly mistaken yes, but I don’t think this applies on this point - he was either truthful or he lied) So I see no motive for lying. No evidence of lying. And any possible question/suggested discrepancy has a simple and prosaic possible explanation.

I can’t say 100% of course, but I’m as near to convinced as can be that Richardson was an honest witness.

(Cue Trevor to tell me that I’m just ‘relying’ on him and that he’s ‘unsafe.’)

[Herlock Sholmes]

Just on the above, and only in light of what this thread discussion is at the moment.

Regarding the accuracy of witnesses testimony solely for the purpose of establishing a definitve t.o.d , I just find it difficult to accept that such testimony is conclusive to that time with this such evidence . imo

Nothing is conclusive Fishy. But it points very heavily and very strongly toward it though, unless we make a conscious and determined effort to try and discredit the witnesses and why should we do that?

[FISHY1118]

[QUOTE=Herlock Sholmes;n792167]

Because the sketch is clearly inaccurate Fishy. The quote that you posted mentions ‘an aperture.’ One gap. If Cadosch hadn’t looked directly or closely at it he wouldn’t have seen through it. He would have passed it in a fraction of a second of course. If he’d seen something he’d have said that he’d seen something. If it was obvious that there couldn’t possibly have been someone in that yard without him clearly seeing them through the fence then the police would have seen this and dismissed his evidence straight away. They didn’t, so clearly it wasn’t obvious.

Im not sure 3 sketchers from 3 different artist can be claimed as 'inaccurate' when they are identical many ways, in relation to the backyard at 29.
The press article confirms the gap as do the sketchers. We dont know if the reporter went to the yard or not ,he may have just stated the gap because it was reported to him that way when in fact the whole fence was gapped .

Im not sure the police at the time had that kind of investigation in mind with any of the witness testimony .

Perhaps if they did, long and richardson , and even perhaps other witnesses in the c5 would have come under much more scrutiny.

[FISHY1118]

Nothing is conclusive Fishy. But it points very heavily and very strongly toward it though, unless we make a conscious and determined effort to try and discredit the witnesses and why should we do that?

Why? . Georges post 1324 whould be a good start.

I think ''Discredit'' is to harsh a word Herlock, i like others are simply trying, by the same process that makes Dr Phillips testimony Shakey , unrieliable ,uncertain, how ever one wishes to phase it , to suggest the same considerstion should be given to all the witneses .

And like i keep saying #post 1324 cant make that point any clearer.

[Herlock Sholmes]

Im pretty sure my post had little to do with him being a liar herlock ,more to the fact i was merely point out that what cadosch heard might not have been the killer and annie conversing in the back yard, and that the noise agaisnt the fence [ yes he was sure about that i agree] might not have been her or the killer either , based on his uncertainty of the 'no' . imo

Fair enough Fishy, if you don’t think that he was lying.

His uncertainty about the ‘no’ wasn’t applied to the noise though. He was certain about it so we have to ask ourselves - if he showed sensible caution on the ‘no’ that appears to show someone that didn’t make unconsidered pronouncements.

If the sound wasn’t connected to Annie’s murder then we have to ask what could it have been. An animal like a cat wouldn’t have made a noise loud enough to have been heard by simply brushing against the fence and it couldn’t have been an innocent person who certainly couldn’t have failed to have seen the body. But you’ve raised a very intriguing possibility Fishy. I don’t think that I has been considered before (as far as I’m aware at least) which is certain to raise the hackles in some quarters.

If Cadosch wasn’t lying and he heard a noise from the yard of number 29 which wasn’t connected to the murder then we have to consider the possibility that not only can we suggest that Chapman wasn’t killed at 4.30 or earlier but she might have died after 5.20! This then brings Elizabeth Long into play. So the suggestion is (and it’s only a suggestion btw)

Richardson sits on the step and the body wasn’t there. Cadosch goes into his yard twice (around 5.20 as a general figure encompassing both visits) and hears someone in the yard of number 29 (perhaps a prostitute and her client, perhaps someone from the house?) That person or persons never come forward because they don’t want to get involved by placing themselves at the crime scene or that they were up to something that they shouldn’t have been? They leave. Long sees Chapman and her killer. They go into the yard and Chapman is killed at around 5.35.

Its only 10 minutes later than if the Cadosch noise was from Annie and her killer and well within the range shown by Jeff’s research.

I can already sense the steam coming out of certain ears. I’d suggest this a possibility nothing more.