The Problem With Times.

[Fleetwood Mac]

If I've missed the answer then I apologise, it's just that when posters think the stated times are reliable, yet the Stride case stands out as one good example against that conclusion. I feel inclined to ask, just how do you think they achieved this 'reliable accuracy', and what apparently went wrong?

I think the answer is that we don't know when witnesses had last looked at a clock, with the exception of Louis.

That makes a world of difference when they estimate a time.

I'm not at all saying they're all reliable. What I'm saying is: don't be so sure the clocks were all miles off.

As an example, publicans could lose their licence in the event they did not know the exact time, driven by a desire to control the supposed indolent and decadent working classes and their drinking hours.

'Just another example of how time and punctuality was very important to the Victorians.

[Wickerman]

In my reading of Baxter's statement, it seems clear that he was interested in establishing a time of death, whether it was his duty to do so or not, and that he found the conflict between the time that Long gave and that of Cadosh to be not a problem. I'll quote a bigger part of the statement below, so we can each interpret it for ourselves.

"…at half-past five, Mrs. Long is in Hanbury-street on her way from home in Church- street, Whitechapel, to Spitalfields Market. She walked on the northern side of the road going westward, and remembers having seen a man and woman standing a few yards from the place where the deceased is afterwards found. And, although she did not know Annie Chapman, she is positive that that woman was deceased. The two were talking loudly, but not sufficiently so to arouse her suspicions that there was anything wrong. Such words as she overheard were not calculated to do so. The laconic inquiry of the man, "Will you?" and the simple assent of the woman, viewed in the light of subsequent events, can be easily translated and explained. Mrs. Long passed on her way, and neither saw nor heard anything more of her, and this is the last time she is known to have been alive. There is some conflict in the evidence about the time at which the deceased was dispatched. 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. It is true that Dr. Phillips thinks that when he saw the body at 6.30 the deceased had been dead at least two hours, but he admits that the coldness of the morning and the great loss of blood may affect his opinion; and if the evidence of the other witnesses be correct, Dr. Phillips has miscalculated the effect of those forces. But many minutes after Mrs. Long passed the man and woman cannot have elapsed before the deceased became a mutilated corpse in the yard of 29, Hanbury-street, close by where she was last seen by any witness."

Thankyou yes, from what I understand the coroner does not enter the time, as in the hour of death, only the date. So, it is not necessary to press Dr Phillips to refine his estimate. Medical knowledge of the period may not be as accurate as today, but they generally did not need an hour of death, except under rare circumstances.
Dr Phillips was sufficiently experienced to know the limitations of the available procedures required with which to estimate a time of death. This he expressed at the inquest, and gave his reasons.
There is every justification for accepting the evidence of Richardson, Cadoche & Long, if we make the required allowances.

[Herlock Sholmes]

The evidence tells us that Chapman died at around 5.25/5.30. Richardson alone proves this. Cadosch and Long serve to firm up the obvious.

[Wickerman]

...
People guessing the time not based on a clock or a watch, aye, in the realms of guesswork. Those who saw a clock, don't be so sure the clock wasn't fairly precise.

If I've missed the answer then I apologise, it's just that when posters think the stated times are reliable, yet the Stride case stands out as one good example against that conclusion. I feel inclined to ask, just how do you think they achieved this 'reliable accuracy', and what apparently went wrong?

[Fleetwood Mac]

I hope you noticed the fact she is reputed to have heard the 5:30 chime 'before' she came to Hanbury street, so while still in Brick Lane (Telegraph), as opposed to having just 'passed' No.29 Hanbury St. (Times). Any difference in time, while not being huge, will be of some significance in any debate concerning the time of the murder.

I don't think it's of any significance, Jon, because poor Annie had long since been carved up.

[Wickerman]

At times Jon, I reckon you place too much emphasis on newspaper discrepancies.

Discrepancies? I fully expected you to be saying "newspaper stories", but yes discrepancies, the papers were full of them in relating crime stories.
I believe it is important to compare press accounts of the same incident to make sure we have as better view of the incident as can be expected - failing a first-hand account of course.
This will naturally involve making comparisons, and any subsequent conclusions may not meet with approval by members who have traditionally bought into one particular version.

As someone who studied history at university, I can say that one of the critical factors in appraising a source is the purpose of the author.

Yes, though a study of contemporary newspapers will hi-lite those papers who take a more liberal view as opposed to conservative outlook on any given incident. There are also a few publications that were viewed to play 'fast & loose' with factual information. One important goal for the editor was to make the story more appealing by adding details not otherwise attested by a witness when relating their experiences.
We must also be careful to recognise when an account is verbatim, or paraphrase. Many times the journalist will shift from first-person to third-person, making it difficult to be sure if the witness was really present or just assumed to be so.

The purpose of newspapers was to get a story out for the readership, i.e. sell newspapers. They're businesses. The exact details of an article is not so important, they're not an investigating organisation: they're in it to sell newspapers.

Correct to a point, they did investigate stories, the editor is the one who will decide what is to be cut or what to include. So, one perfect example is the press account of an inquest. We might have half-a-dozen versions of one witnesses testimony, and all of them different. This is the editor choosing what he thinks is relevant.
It has become habitual by some members to accept and only repeat their one preferred version, without checking & taking into account what other newspapers attribute to the same witness. adding detail in order to create excitement for the reader is to be expected. Whether this added detail is always factual is to be questioned.
All of this is, as you say, towards one end - to sell newspapers.

I reckon you can take it as fact that newspapers repeated second hand stories (from other sources) at times and the exact details weren't so important, and you can take it as read that first hand accounts didn't contain the exact details verbatim.

The weekend papers (Reynolds/Lloyds/etc.) were notorious for copying the dailies. We rarely find any new information in a weekend newspaper - mostly they are summaries of the weeks events concerning the Whitechapel murders. The evening papers (Evening News/Star/Echo/etc.) quite often copied the morning papers. In some cases they will use a whole paragraph, then insert a summary of their own creation where the detail tends to tail off. The evening papers went to press about 3:00 pm so we are only able to read half the proceedings of an inquest in an evening paper, the reporter had to leave before the inquest terminated in order to meet the publication times.

Morning papers are typically the preferred source, but whatever happened in an ordinary day is already a day late by the time it is published, as the dailies go to press overnight. Then there are the Agencies who sold stories to newspapers by telegraph (Press Association/Central News/Telegraph, etc). Some dailies will buy a story, then edit it down, and/or add content of their own. Not always giving credit to the agency for the story.
One example from the Evening News begins "Our representative interviewed the witness xxxxxxxx today, etc" but if we look around we see the same story published in another newspaper that reads: "The Press Association say, our representative interviewed the witness xxxxxxx today, etc.", so the Evening News took credit for that story, so as well as selling copy, it's also about "oneupmanship".
It's true, you have to be really careful in dealing with newspapers, you have to do your homework, they cannot be taken at face value, and you have to compare as many versions of the same story across different newspapers.

As I say, they're not investigative organisations. It's a case of getting the story out to sell newspapers and the exact details are not of much consequence to them.

Well, we have it from the police (Warren?), that these journalists are perpetually following detectives around to re-interview witnesses visited by detectives. We also have instances where the press have investigated a story, but for the most part it is true they are more interested in capturing a story than checking for accuracy, or at least that is the way it seems.

On this occasion, however, I agree. Elizabeth didn't suggest it was precisely half 5. In fact, in both sources she clearly states: "about".

What we do know from all accounts, is that she heard the clock strike half 5. It follows that when Elizabeth stated about half 5, the reasonable conclusion is that her 'about' was a minute or two after half 5. Clearly she couldn't have meant before half 5 given that she heard the clock strike half 5, and she didn't say half 5, so about half 5 can only reasonably mean a minute or two afterwards.

I hope you noticed the fact she is reputed to have heard the 5:30 chime 'before' she came to Hanbury street, so while still in Brick Lane (Telegraph), as opposed to having just 'passed' No.29 Hanbury St. (Times). Any difference in time, while not being huge, will be of some significance in any debate concerning the time of the murder.

[Fleetwood Mac]

Staying with Mrs Long, even her assumption is open to debate, likely not any of her doing, fault could lay with the press.

You'll notice in the Daily Telegraph she is reputed to say:
"I knew the time, because I heard the brewer's clock strike half-past five just before I got to the street".

Just before she got to Hanbury Street.

Whereas in the Times, we read:
"It was about 5:30. She was certain of the time, as the brewers' clock had just struck that time when she passed 29, Hanbury-street."

When she passed No.29.

There's likely a couple of minutes difference depending on what "just before" means, compared with 'just passing".

At times Jon, I reckon you place too much emphasis on newspaper discrepancies.

As someone who studied history at university, I can say that one of the critical factors in appraising a source is the purpose of the author.

The purpose of newspapers was to get a story out for the readership, i.e. sell newspapers. They're businesses. The exact details of an article is not so important, they're not an investigating organisation: they're in it to sell newspapers. I reckon you can take it as fact that newspapers repeated second hand stories (from other sources) at times and the exact details weren't so important, and you can take it as read that first hand accounts didn't contain the exact details verbatim. As I say, they're not investigative organisations. It's a case of getting the story out to sell newspapers and the exact details are not of much consequence to them.

On this occasion, however, I agree. Elizabeth didn't suggest it was precisely half 5. In fact, in both sources she clearly states: "about".

What we do know from all accounts, is that she heard the clock strike half 5. It follows that when Elizabeth stated about half 5, the reasonable conclusion is that her 'about' was a minute or two after half 5. Clearly she couldn't have meant before half 5 given that she heard the clock strike half 5, and she didn't say half 5, so about half 5 can only reasonably mean a minute or two afterwards.

[Herlock Sholmes]

Hi Herlock,

I had come across a paper that had some relevant bits, but it didn't have empirical data of the sort we could translate into estimates of the margins of error we should be considering unfortunately. However, it did reference such things as letters of complaint sent to the news with regards to how frustrating it was when every clock read a different time, and that London should do something about it. One letter mentions how in some cities, Paris I think, one could listen to the clocks chime the hour for 30 minutes, and that London wasn't all that much better.

While the train stations did synchronize to GMT, by telegraph I think, for a few decades (a bit earlier than the 1880s), there was the situation where a town had the "local time" and the trains had GMT, so local clocks would read a different time than the train station clocks! There were people who would get the "official time", and then go around to different businesses and people and update them on the correct time (for a fee; sort of like getting the newspaper delivered to your house when that was a thing). So there was some interest in standardizing time, the city itself did not actually make a great effort to ensure its own clocks were correct (politicians like to set the rules, not follow them after all ).

So really, there is no sound argument for considering the times that witnesses give should align (even if two people had access to the same clock, unless they both read it at the time of the event they are giving a time for, like Dr. Blackwell who read his watch when he arrived at the scene), we end up with the possibility one, the other, or both misrecall the time the clock said when the looked at it, and/or misestimate the duration of time that had passed between having looked at the clock and the event to which they are testifying. Once two witnesses are likely to have "set their time" by different clocks, we add a third source of error, which is the clock-sync error.

What I think sometimes people think is that because, of course, it is possible for two clocks to have been in sync, how do we know they weren't? Obviously, we don't know by how much two clocks differ. We have no information about the specific clocks in question (often we don't even know for sure which clock people used, although for some we do). What margins of error are all about is how to view the information under exactly these conditions - when you don't know, you have to consider the margins of error. If we did know (i.e. the police compared two clocks, realised one was faster by 5 minutes, then we can adjust stated times to be "in sync", we no longer need margins of errors for those two clocks). Margins of error are what we use when we do not have the information - to just pick a possibility and argue that must be the case is indefensible, it's a guess, and in the end, it gets you nowhere in solving something although it may get you where you want to be. (Note, this is not quite the same thing as considering one possibility, and then comparing it to when you pick another; then you are looking at the consequences of the two choices - if they both lead to the same conclusion, then you can argue it doesn't matter which was the case, both lead to the same conclusion. If they lead to different conclusions, then you are stuck with not knowing which branch to follow, but you do know what information you need to find to make the decision. If that information is how much two clocks differed, well, sadly, that information is no longer available, and we're stuck).

- Jeff

Hi Jeff,

Well put, as ever. Yes it’s when we’re dealing with a lack of certainty that margins for error provide a range to work with and guard us against unfounded confidence (or at least they should do.) Perhaps we don’t always realise how much of the time we’re working in the dark in our investigations of crimes that happened so long ago? Virtually everything throws up a list of questions that we have no way of answering with confidence. In the case of Heschberg for example:

1. How did he arrive at his estimated time?
2. Was it from someone else?
3. How accurate was the other persons time if that was the case?
4. Was it from the club’s clock?
5. How accurate was that clock?
6. How long before the body was found did he look at the clock?
7. How long was the gap between seeing the clock and being interviewed?
8. How good was he at estimating periods of time?

We can’t answer any of those questions. So how can his hardly confidant “around 12.45 I should think,” be considered accurate enough to uphold a theory? No one is saying that he couldn’t have been accurate of course, just that there is a very real possibility that he wasn’t. And when we look at the witnesses as a whole we only have him and Kosebrodsky who point to an earlier ToD. So the likelihood has to be that they were mistaken. Or else we would have to accept as a fact that 2 concurring witnesses can’t both be wrong.

[JeffHamm]

Thanks for posting this info again Jeff. Didn’t you also post some evidence on timekeeping/clocks in the Victorian era too at one point?

Hi Herlock,

I had come across a paper that had some relevant bits, but it didn't have empirical data of the sort we could translate into estimates of the margins of error we should be considering unfortunately. However, it did reference such things as letters of complaint sent to the news with regards to how frustrating it was when every clock read a different time, and that London should do something about it. One letter mentions how in some cities, Paris I think, one could listen to the clocks chime the hour for 30 minutes, and that London wasn't all that much better.

While the train stations did synchronize to GMT, by telegraph I think, for a few decades (a bit earlier than the 1880s), there was the situation where a town had the "local time" and the trains had GMT, so local clocks would read a different time than the train station clocks! There were people who would get the "official time", and then go around to different businesses and people and update them on the correct time (for a fee; sort of like getting the newspaper delivered to your house when that was a thing). So there was some interest in standardizing time, the city itself did not actually make a great effort to ensure its own clocks were correct (politicians like to set the rules, not follow them after all ).

So really, there is no sound argument for considering the times that witnesses give should align (even if two people had access to the same clock, unless they both read it at the time of the event they are giving a time for, like Dr. Blackwell who read his watch when he arrived at the scene), we end up with the possibility one, the other, or both misrecall the time the clock said when the looked at it, and/or misestimate the duration of time that had passed between having looked at the clock and the event to which they are testifying. Once two witnesses are likely to have "set their time" by different clocks, we add a third source of error, which is the clock-sync error.

What I think sometimes people think is that because, of course, it is possible for two clocks to have been in sync, how do we know they weren't? Obviously, we don't know by how much two clocks differ. We have no information about the specific clocks in question (often we don't even know for sure which clock people used, although for some we do). What margins of error are all about is how to view the information under exactly these conditions - when you don't know, you have to consider the margins of error. If we did know (i.e. the police compared two clocks, realised one was faster by 5 minutes, then we can adjust stated times to be "in sync", we no longer need margins of errors for those two clocks). Margins of error are what we use when we do not have the information - to just pick a possibility and argue that must be the case is indefensible, it's a guess, and in the end, it gets you nowhere in solving something although it may get you where you want to be. (Note, this is not quite the same thing as considering one possibility, and then comparing it to when you pick another; then you are looking at the consequences of the two choices - if they both lead to the same conclusion, then you can argue it doesn't matter which was the case, both lead to the same conclusion. If they lead to different conclusions, then you are stuck with not knowing which branch to follow, but you do know what information you need to find to make the decision. If that information is how much two clocks differed, well, sadly, that information is no longer available, and we're stuck).

- Jeff

[Wickerman]

To use Elizabeth Long as an example, when she said she knew the time according to a clock, you cannot reasonably bend it 10 minutes on the back of a broad sweep: "the clocks were awful in a slum in Victorian London", and of course, in the event you're going to go down that road then you'd have to argue that she could equally have been wrong 10 minutes the other way.

Staying with Mrs Long, even her assumption is open to debate, likely not any of her doing, fault could lay with the press.

You'll notice in the Daily Telegraph she is reputed to say:
"I knew the time, because I heard the brewer's clock strike half-past five just before I got to the street".

Just before she got to Hanbury Street.

Whereas in the Times, we read:
"It was about 5:30. She was certain of the time, as the brewers' clock had just struck that time when she passed 29, Hanbury-street."

When she passed No.29.

There's likely a couple of minutes difference depending on what "just before" means, compared with 'just passing".

[Lewis C]

I was addressing what you said - "The more difficult question is, How much of an allowance?".

I was giving the reason why Baxter said "not very great, or very important".

In my reading of Baxter's statement, it seems clear that he was interested in establishing a time of death, whether it was his duty to do so or not, and that he found the conflict between the time that Long gave and that of Cadosh to be not a problem. I'll quote a bigger part of the statement below, so we can each interpret it for ourselves.

"…at half-past five, Mrs. Long is in Hanbury-street on her way from home in Church- street, Whitechapel, to Spitalfields Market. She walked on the northern side of the road going westward, and remembers having seen a man and woman standing a few yards from the place where the deceased is afterwards found. And, although she did not know Annie Chapman, she is positive that that woman was deceased. The two were talking loudly, but not sufficiently so to arouse her suspicions that there was anything wrong. Such words as she overheard were not calculated to do so. The laconic inquiry of the man, "Will you?" and the simple assent of the woman, viewed in the light of subsequent events, can be easily translated and explained. Mrs. Long passed on her way, and neither saw nor heard anything more of her, and this is the last time she is known to have been alive. There is some conflict in the evidence about the time at which the deceased was dispatched. 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. It is true that Dr. Phillips thinks that when he saw the body at 6.30 the deceased had been dead at least two hours, but he admits that the coldness of the morning and the great loss of blood may affect his opinion; and if the evidence of the other witnesses be correct, Dr. Phillips has miscalculated the effect of those forces. But many minutes after Mrs. Long passed the man and woman cannot have elapsed before the deceased became a mutilated corpse in the yard of 29, Hanbury-street, close by where she was last seen by any witness."

[Herlock Sholmes]

I've done you the courtesy of short-handing your OP, Herlock.

These are your points in condensed form:

1) "Poorly synchronised clocks and watches" (even today). You expect said "poor synchronisation" to be "much greater" in "Victorian London" and in "a slum where many didn't own a watch or clock".

2) "We have to make allowances for the fallibility of memory when witnesses are estimating periods of time".

The above is the meat of your OP, and of course you go on to apply these pertinent points to a theory that you hold.

What theory? I have no theory? I’ve been debating with a person that certainly does have a theory and who relies on spot-on times to bolster it.

On point 2, absolutely. This stands for any case in any age. But, don't we all know this? Doesn't any and every investigating officer know this? In the event we have a witness who says about quarter past 1, and we don't know whether or not that person has looked at a clock for the preceding couple of hours, then of course caution should be exercised; and that applies in any case in any age.

Apparently we don’t all know this. I made this poll because Michael continually tries to suggest that we should take times as they are stated in testimony. Andrew appears to agree with him. Thankfully they are in a minority.

On point 1, I think you're too quick to rush in to the 'we can't trust the times' hypotheses.

1) The Victorians were sticklers for punctuality. A lot more so than we are. That was part of their culture. At our work, we're judged on getting the job done and so a couple of minutes turning up late here and there isn't much of an issue. In their age, good timekeeping was a mark of character.

Im not sure about that at all. Maybe in office work but not in factory work. I spent most of my working life as a supervisor and manager and non-staff employees were allowed two minutes past clocking in time. Three minutes went on their record as ‘late.’ A very few lates over a period meant the disciplinary procedure. I’ve known quite a few people sacked for regular lateness of one or two minutes.

2) They had a master clock from which they referenced time and set their clocks, just as we do.

Maybe but we check ours online or from the TV or radio. They checked against other clocks which they had to trust. Also, how at any given time can we be sure when a clock was last put right?

3) The master clocks needed to be in good from, otherwise you had trains smashing into one another where and when they shared the same track.

Ive no doubt about railway company clocks but what about the clock in the shop window that Diemschitz saw? Or a clock that Fanny Mortimer might have had?

4) Nobody at any inquest suggested the times couldn't be trusted. In fact, witnesses were pressed for times. In the event the clocks couldn't be trusted they'd have known through the experience of turning up for a train only to find the train had gone and they were 10 minutes late.

You appear to be treating slum dwellers as if they were city gents off to the Stock Exchange.

To use Elizabeth Long as an example, when she said she knew the time according to a clock, you cannot reasonably bend it 10 minutes on the back of a broad sweep: "the clocks were awful in a slum in Victorian London", and of course, in the event you're going to go down that road then you'd have to argue that she could equally have been wrong 10 minutes the other way.

And I’ve said numerous times that we can’t have a margin for error going just one way. A margin for error both ways has to be applied and we should prove or eliminate something by refusing to acknowledge it.

I re-post the piece originally posted by George from a man with a far greater knowledge of watches/clocks and timekeeping than any of us:

'The issue is I believe of how accurate was time measured in London in the 1880s.

I think it depends who you were. If you were well off you had a good pocket watch, the clocks in your house were wound weekly by a local clockmaker, who one would hope had a reasonable standard of time. Any decent clockmaker would have a shop regulator, a clock that told time accurately; let us say it might vary by 10 seconds a week at the most. Then there is the problem of how the regulator was set. A sundial was one possibility, a good large sundial could tell the time to a minute, but the user needs to be aware of the equation of time; that difference between solar time and mean time. The striking of Big Ben is accurate to a second... you have to allow for the speed of sound so a map was published for London showing the delay. Even so, you are talking about 10 seconds at the most, and if you know where you are, then you can compensate. London then was not so built up, so I think the clock could be heard over a larger range. A new St Paul's Cathedral clock was installed in 1892, the previous clock was by Bradley and was set by a sundial, it could be probably read to within 20 seconds. Several balls were erected in London from 1852 onwards, these were released by telegraphed time. Time signals could be telegraphed to those who paid for them. The British Horological Institute (BHI) installed a telegraph instrument in 1862, and this was used to correct a regulator. Members could then come with an accurate watch to 'get the time'. Ruth Belville took a pocket chronometer round Clerkenwell, I think here customers were mainly chronometer makers. Pretty crazy since they could get the time by telegraph if they paid, or from the BHI in Northampton Square. So perhaps the 'Gentleman' in the street in London would know the time by his watch to within half a minute. This being derived from a clock at home and that being set by his clockwinder. Taking the traders, they would have watches, but probably no clockwinder to regulate their house clocks. So they are down to using Big Ben, St Paul's, the time ball for accurate time and the local church clock otherwise. The working class generally did not have watches, though they possibly had cheap clocks that were being imported from America. Watches could be had from America and Switzerland for a tenth of the price of an English watch. You will have heard probably of the dollar watch. Their watches and clocks would have been set by the nearest public clock. I remember an old man telling me that during WW1 when he was a lad, his mother asked him to run down to the church and get the time. So we have an accuracy of several minutes. This is somewhat confused by the local church clock. Certainly in village we know that early in the 19th century variations from village to village were of the order of plus and minus 20 minutes or even more. I am not familiar with church clocks in Whitechapel, but their clock setting was often at the whim of the clock winder. Perhaps plus or minus 10 minutes could be expected, or even more. So, if the working man's watch or clock was within 10 minutes of true time, I would think that good for the 1880s. You need to understand why time was needed. First to regulate a local community. The absolute standard did not really matter, that all met together at the appointed time was sufficient. In larger communities the railway was the one that called the shots, the train left at Railway (Greenwich) time and it was tough if you missed it due to your poor clock or watch. Factories were another corporate gathering place. Again, their standard was absolute to them, so provided all were through the gate by the set time on their clock, then all was OK. The fact that it was fast or slow of GMT was irrelevant. Churches publicised their services by bells, so the exact time of starting was not too important.

Overall I think that if you found a clock in the East End that was telling time to within 10 mins of GMT you were doing well.

Regards, Chris McKay

[Fleetwood Mac]

That's not what I claimed at all.

I've done you the courtesy of short-handing your OP, Herlock.

These are your points in condensed form:

1) "Poorly synchronised clocks and watches" (even today). You expect said "poor synchronisation" to be "much greater" in "Victorian London" and in "a slum where many didn't own a watch or clock".

2) "We have to make allowances for the fallibility of memory when witnesses are estimating periods of time".

The above is the meat of your OP, and of course you go on to apply these pertinent points to a theory that you hold.

On point 2, absolutely. This stands for any case in any age. But, don't we all know this? Doesn't any and every investigating officer know this? In the event we have a witness who says about quarter past 1, and we don't know whether or not that person has looked at a clock for the preceding couple of hours, then of course caution should be exercised; and that applies in any case in any age.

On point 1, I think you're too quick to rush in to the 'we can't trust the times' hypotheses.

1) The Victorians were sticklers for punctuality. A lot more so than we are. That was part of their culture. At our work, we're judged on getting the job done and so a couple of minutes turning up late here and there isn't much of an issue. In their age, good timekeeping was a mark of character.

2) They had a master clock from which they referenced time and set their clocks, just as we do.

3) The master clocks needed to be in good from, otherwise you had trains smashing into one another where and when they shared the same track.

4) Nobody at any inquest suggested the times couldn't be trusted. In fact, witnesses were pressed for times. In the event the clocks couldn't be trusted they'd have known through the experience of turning up for a train only to find the train had gone and they were 10 minutes late.

To use Elizabeth Long as an example, when she said she knew the time according to a clock, you cannot reasonably bend it 10 minutes on the back of a broad sweep: "the clocks were awful in a slum in Victorian London", and of course, in the event you're going to go down that road then you'd have to argue that she could equally have been wrong 10 minutes the other way.

[Herlock Sholmes]

A good post Jeff.

An easy test at how poor people are at estimating time is is to estimate one minute. Wait until a clock ticks over to a minute and then look away without counting and then check the clock after you think a minute has elapsed. I suggest that the result would be even worse when people are not consciously trying to calculate the time which has elapsed.

I'll be interested to hear the results if anyone tests themselves.

Regards, Gazza

I just tried that twice Gazza. Once without counting the seconds in my head and once with counting.

Without counting I got 46 seconds. (Perhaps to be expected from a man with zero patience)

With counting I got 53 seconds.

It’s just not difficult to see how someone like Heschberg might have seen the clock in the club at some point before hearing the alarm, let’s say 12.15 for discussions sake. Then he’s interviewed by the police but when? They would have interviewed all members and staff so it might been, again for discussions sake about 2.30. So he has to remember seeing the clock 2 hours ago and recall the time then estimate the time period between then and when he heard to alarm. Can anyone really have a level of confidence that he couldn’t have been mistaken? Surely the more likely explanation about Heschberg and Kozebrodsky (and 50% of Spooner’s testimony) is that they were simply mistaken considering they disagree with the majority of witnesses?

[Herlock Sholmes]

Actually, no, we're rubbish at it. I presented actual research into this very point, our ability to estimate the duration of a time interval, and we're pretty awful at it when looking at how wide the margins of error are. The short version is that people tend to overestimate short durations (up to about an hour; so if they say it was 5 minutes, it was probably more like 3 minutes, type thing). After an hour or so, we tend to underestimate the duration (so if we say it was 1.5 hours, it was probably more like 1.7 hours type thing - by the way, those aren't the actual numbers from the study, I'm just using them to clarify what I mean).

Here's a table I put together from the research, which shows the relationship between a stated duration (the leftmost column) and actual durations showing the lower (min), average, and upper (max) values for the actual durations of the event. By lower and upper, those combined comprise the 95% confidence interval. For example, when someone says an interval was 5 minutes, then the actual interval could be somewhere between 1m 38s and 15m 06s, with the average actual interval being around 3m 37s (we overestimate short intervals).

Of course, some people may be better at this than others, so maybe you're particularly good at it. But anecdotal evidence doesn't help us, because we have no idea of how good or bad the witnesses were, so we need to recognize that duration estimation by people is, in the end, not all that impressive.



- Jeff

Thanks for posting this info again Jeff. Didn’t you also post some evidence on timekeeping/clocks in the Victorian era too at one point?