Witness Testimony: Albert Cadosche

[Herlock Sholmes]

At the Eddowes inquest:

Constable Long

[Coroner] Had you been past that spot previously to your discovering the apron? - I passed about twenty minutes past two o'clock.

Detective Halse

I came through Goulston-street about twenty minutes past two,

Do we consider this to be a problematic conflict?

Hypothetical Scenario.

There’s a fight in a street. Two days later one of the men dies as a result of his injuries. Two weeks later two witnesses came forward both of whom had gone on holiday abroad after seeing the fight so were unaware of the significance of their evidence. One witness says that she’d seen the fight at 1.50 pm and that she’d arrived at her time by the fact that she’d left her house at around 1.45 and the location was around 5 minutes away. The other witness said that he’d seen the fight a little after 1.30 pm and he took his time because he’d left his phone in his car and so had asked a passerby who told him that it was 1.30 around 2 minutes or so before he saw the fight.

So should the police doubt the validity of these two witnesses because there is an apparent 20 minute discrepancy between the two?

[JeffHamm]

That is wrong.

One cannot conclude that there is no conflict on the basis of the twin assumptions that one clock was at least six minutes slow and the other clock was at least six minutes fast (but not vice-versa).

That is not evidence.

Which is why I didn't say you could conclude there wasn't a conflict.

I said you cannot conclude there is a conflict. That's not the same thing.

But unless you can conclude there is a conflict, then there is no conflict to resolve. That's the problem. You have to be able to show the evidence is conflicting - you have to be able to conclude there is a conflict. It's trivial to find combinations from ranges that don't work, but if the ranges allow for a resolution, then the information we have cannot be shown to be in conflict. That is why it is an error to say that Long and Cadosche's testimony conflict, because they cannot be shown to conflict due to the imprecision of the information we have to work with.

That isn't saying we are able to state with certainty what the times were, we obviously can't. If we could we wouldn't be dealing with ranges we would have the actual information about what the two clock's synchronisations were, and at that point we would be able to know whether or not there was or was not a conflict in the stated times. (Note, if we had that information, we might still have to consider the possibility that Long has misrecalled what chimes she had heard, and to resolve that, we would have to be able to interview her, find out where she lived, be able to work out what time she left home, and track her movements, find out how fast she walks, and so forth, and from that work out what time it would be when she passed the Brewer's clock - again, all things we cannot do, so we are left with having to consider a range of memory accuracies - from she recalls the chimes correctly to she recalls the chimes incorrectly - and we cannot resolve that because there is no longer any way to get the specific details. We can try to estimate the time, but her actual residence seems to be unsure. Having examined the distance from various locations people have suggested for her residence, though, the distance from any of them would suggest she should have passed the Brewer's clock around 5:15, making a misrecall of the chime something one has to consider; well I suppose one doesn't have to, but certainly one should.)

But back to clock sync only, we don't have the information about the specific clock sync's, so we are left with ranges to deal with, and the ranges we have do not demonstrate a conflict, so we cannot say that their testimony's are in conflict and so we cannot say there is a conflict to resolve.

- Jeff

[PRIVATE INVESTIGATOR 1]

However, when faced with unknown actual values, then as long as the ranges allow for a resolution one cannot conclude there is a conflict.

That is wrong.

One cannot conclude that there is no conflict on the basis of the twin assumptions that one clock was at least six minutes slow and the other clock was at least six minutes fast (but not vice-versa).

That is not evidence.

[JeffHamm]

As I explained in post #698, any time you're dealing with ranges of times for 2 events, it's always possible to pick times within those ranges whereby the events cannot be reconciled, and clearly it would be nonsense to say that nothing can be reconciled with anything else. To establish that there's no conflict between 2 testimonies, it is only necessary to find one way within the ranges in which the testimonies don't conflict.

That's exactly it! Obviously, if it were possible, one would then test that possibility by going to the two clocks and working out what the relative time differences are between them. But at that point, one no longer needs to deal with ranges because by measuring the actual clock settings one no longer has to deal with a range of possibilities. However, when faced with unknown actual values, then as long as the ranges allow for a resolution one cannot conclude there is a conflict. As such, with regards to Long and Cadosche, we cannot conclude there is a conflict, and therefore given the information we have there is nothing to resolve.

That is how range information is interpreted. It's not an opinion, it is simply the method by which one analyses such noisy data. In an active police investigation today, of course, such noisy information is unacceptable and one would be sent out to clarify the settings on the clocks. That wasn't something the police did at the time, but methods have improved since then.

- Jeff

[PRIVATE INVESTIGATOR 1]

To establish that there's no conflict between 2 testimonies, it is only necessary to find one way within the ranges in which the testimonies don't conflict.

This is our basic disagreement.

The fact that if certain conditions applied, the conflict between two testimonies would be resolved, does not of itself resolve the conflict!

[Lewis C]

It is being suggested that the Brewery's clock was fast and the church's clock slow, in both cases by five minutes.

Even if one assumes that, according to their testimonies, Long passed the couple at 5.31 and Cadoche heard 'no' at 5.22, Chapman would have had one minute in which to finish her conversation, walk to number 29, walk through the passage, open the door to the yard, enter the yard and say 'no'.

If the Brewery's clock was slow and the church's clock fast, then the two testimonies cannot be reconciled.

If the Brewery's clock was fast and the church's clock slow, in both cases by four minutes, then the two testimonies cannot be reconciled.

I suggest that the Brewery's clock would have had to be fast and the church's clock slow, in both cases by at least six minutes.

I suggest also that that is nothing more than an assumption.

As I explained in post #698, any time you're dealing with ranges of times for 2 events, it's always possible to pick times within those ranges whereby the events cannot be reconciled, and clearly it would be nonsense to say that nothing can be reconciled with anything else. To establish that there's no conflict between 2 testimonies, it is only necessary to find one way within the ranges in which the testimonies don't conflict.

[PRIVATE INVESTIGATOR 1]

It is being suggested that the Brewery's clock was fast and the church's clock slow, in both cases by five minutes.

Even if one assumes that, according to their testimonies, Long passed the couple at 5.31 and Cadoche heard 'no' at 5.22, Chapman would have had one minute in which to finish her conversation, walk to number 29, walk through the passage, open the door to the yard, enter the yard and say 'no'.

If the Brewery's clock was slow and the church's clock fast, then the two testimonies cannot be reconciled.

If the Brewery's clock was fast and the church's clock slow, in both cases by four minutes, then the two testimonies cannot be reconciled.

I suggest that the Brewery's clock would have had to be fast and the church's clock slow, in both cases by at least six minutes.

I suggest also that that is nothing more than an assumption.

[Lewis C]

But is it not reasonable to assume that people were in the habit of arriving at work at about the same time each day?

For example, Lechmere said that he was in danger of arriving late and that he started work at 4 a.m.

For some jobs, a few minutes wouldn't matter. Mreover, in Long's case, the possibility being raised is that maybe she got to work earlier, which certainly wouldn't be a problem. It might be the case that Cadosch wouldn't really have been late in either case.

[PRIVATE INVESTIGATOR 1]

If there's evidence that Cadosch and Long were required to be at work at precise times, I'm not aware of it.

But is it not reasonable to assume that people were in the habit of arriving at work at about the same time each day?

For example, Lechmere said that he was in danger of arriving late and that he started work at 4 a.m.

[Lewis C]

But is it reasonable to assume that Long could not tell the quarter past and half past chimes apart, or that she did not realise that she arrived at work early, or that Cadoche did not realise that he arrived at work late, and that his colleagues and employees did not realise it either?

First, with Long, I'm only saying that more than one thing is possible. Even if she could tell the 2 chimes apart, maybe she wasn't making a point of noticing it at the time, or maybe she misremembered it later. If there's evidence that Cadosch and Long were required to be at work at precise times, I'm not aware of it.

[Herlock Sholmes]

On the subject of the gap of time between the ‘no’ and the noise then 5 minutes would be an issue as the killer clearly wouldn’t have held Annie up for 5 minutes before allowing her to fall to the ground.


The problem with this point of course is that it requires to unsafe assumptions at both ends to make this a valid point. The word ‘no’ might or might not have been the commencement of the attack; we have no way of knowing. It could equally have been a response to a question. In making a suggestion I’m only seeking to provide an example. I’m claiming nothing as a fact…just to be clear. What if the killer, on hearing Cadosch, asked if they should find another spot. To which Annie replied “NO…he’s gone inside now so he’s not going to come back out.” Other suggestions could be made of course but the point is that we can’t know the context of the ‘no.’


The second assumption is that the noise was Annie’s body falling against the fence. But this could have course been made by the killer himself bumping against the fence.


So the situation can be made to appear unlikely if we accept two unsafe assumptions. Surely we are better served by acknowledging all possibilities without assuming that we know exactly what went on. The gap between the ‘no’ and the noise therefore should not be viewed as detrimental to any suggestion.

[Herlock Sholmes]

Firstly, there is no need to make an assumption involving 15 minutes in regard to either Long or Cadosch’s evidence. It doesn’t matter how the coroner tried to reconcile the two times because that was simply the only suggestion considered at that time. Now, after additional thought and consideration we can now see that the only margin for error required is one of around 5 or 6 minutes. We recall again what Cadosch said at the inquest:


“I got up about a quarter past five in the morning, and went into the yard. It was then about twenty minutes past five, I should think.”


So it couldn’t be clearer that there was a level of uncertainty. He was uncertain about exactly what time he got up and he was uncertain about the length of time between him getting up and going out.

So this is quite clear…..Cadosch was estimating and was uncertain about these times. We can also add of course that he makes no mention of how long he was in the toilet on both visits. He also estimates that it was 3 or 4 minutes later when he made his second visit….so more uncertainty.

So we cannot dispute the fact that these times are uncertain estimates which we have no way of verifying.


Secondly, just as we know that clocks and watches aren’t perfectly accurate and synchronised even in 2023 with the massive advancements in technology it should be obvious that we have to allow for the same in 1888 (135 years ago) Not to do this across the board would be to risk getting a false picture by dismissing or assuming things unnecessarily. Whichever side of any debate that we might find ourselves on.


Thirdly, Mrs Long said:


“On Saturday, Sept. 8, about half past five o'clock in the morning, I was passing down Hanbury-street, from home, on my way to Spitalfields Market. I knew the time, because I heard the brewer's clock strike half-past five just before I got to the street.”


So, according to Mr. Long it was probably still 5.30 by the brewer’s clock as she passed the couple.


————


Now, let’s dismiss the perhaps unlikely sounding 15 minutes and talk of a mere 5 or 6 minutes. It’s difficult to conceive of anyone claiming that it would have been unlikely for a clock to have been 5 or 6 minutes out in 1888. As we speak, I can find two clocks in my own house that are out by at least that much. So a 5 or 6 minutes discrepancy is absolutely nothing. Perfectly normal, everyday and entirely reasonable.

So the suggestion is that when Long passed the couple it was actually 5.25 because the brewers clock was 5 minutes fast (as clocks often are)


And if it’s not unreasonable to suggest that the brewers clock might have been 5 minutes fast (or maybe more?) then it cannot be unreasonable to suggest that when Cadosch estimated that he’d got up at about 5.15 then he could very easily have been 5 minutes out and that it was actually 5.20. Or maybe it was 5.18 and the estimated gap between him getting up and going into the yard was slightly longer. No matter which way, the margin for error is tiny. But small periods can be significant.


And so with just the merest margin for error we have Long and Cadosch tying up perfectly. No need for an unlikely sounding 15 minutes. And certainly no need for any suggestion of manipulating a time to suit. I believe that these margins for error should be applied across the board to every single time quoted in this case. No exceptions whatsoever ever so complete fairness.



So there are clearly no grounds for dismissing Long or Cadosch on the time. No modern police investigation (if they had no way of testing the accuracy of clocks at the time of the crime) would dismiss anything on the bases of 5 minutes or so.

[PRIVATE INVESTIGATOR 1]

If there's a way that the testimonies can reasonably be reconciled, then there's no conflict.

Here is how the coroner tried to reconcile the two witnesses' testimonies:

There is some conflict in the evidence about the time at which the deceased was despatched. It is not unusual to find inaccuracy in such details, but this variation is not very great or very important. She was found dead about six o'clock. She was not in the yard when Richardson was there at 4.50 a.m. She was talking outside the house at half-past five when Mrs. Long passed them. Cadosh says it was about 5.20 when he was in the backyard of the adjoining house, and heard a voice say "No," and three or four minutes afterwards a fall against the fence; but if he is out of his reckoning but a quarter of an hour, the discrepancy in the evidence of fact vanishes, and he may be mistaken, for he admits that he did not get up till a quarter past five, and that it was after the half-hour when he passed Spitalfields clock.

Do you find that convincing?

Cadoche gets up and thinks it is about 1/4 of an hour earlier than it really is, the Spitalfields church clock is about 1/4 of an hour earlier than it really is, and when Cadoche arrives at work about a quarter of an hour late, no-one notices?


P.S. the coroner underestimated the time interval between the two sounds.

[PRIVATE INVESTIGATOR 1]

In terms of the balance of probability, it's difficult to say and cannot be quantified, but I reckon there is enough there to suggest that Albert's witness statement is not cast-iron and is open to a reasonable degree of doubt (in terms of what he recollected versus that which actually happened).

What is the probability of the two sounds he heard being connected when they occurred about five minutes apart?

If Chapman and her murderer were having a conversation, during which she said 'no', and she fell against the fence about five minutes later, what is the murderer supposed to have been doing during the intervening five minutes?

[PRIVATE INVESTIGATOR 1]

The question at hand is whether there's a conflict between the testimonies of Cadosch and Long. If there's a way that the testimonies can reasonably be reconciled, then there's no conflict.

But is it reasonable to assume that Long could not tell the quarter past and half past chimes apart, or that she did not realise that she arrived at work early, or that Cadoche did not realise that he arrived at work late, and that his colleagues and employees did not realise it either?