Further to that, another thing that frequently happens (in everyday life, as well as in witness testimony, where it is a regular problem) is conflation of different, related memories.
To illustrate, let me refer to a different unsolved crime. When little Christine Jessop disappeared from Queensville in 1984 (story here for anyone who is interested and/or doesn't remember the details:
http://www.cbc.ca/news/canada/toron...l-hoping-for-justice-30-years-later-1.2784486 ) , the initial investigation was, by today's standards, not well done. But, neighbours, townspeople, playmates etc. were interviewed, and a tentative outline of Christine's movements that afternoon was pieced together.
As in most cases, people came forward after the publicity about the missing child to share what they knew, or thought they knew. But there were inconsistencies: reports that Christine bought bubble gum at the store at a certain time; a driver at the light who was sure he saw Christine talking to an older boy (a teenager) on the corner; a couple who thought they saw a child struggling with a driver in a car, and who followed that car for a few blocks because they were suspicious, and so on. Trouble was, in some cases, it appeared (in hindsight) that people had conflated real memories of seeing Christine, or a child similar, but had the day wrong (in some cases) or the time of day, in others. There were similar problems with witness evidence trying to pin down what happened to little Nicole Morin, who went missing in 1985 from her apartment building on the West Mall. It's seems fairly certain that some of the accounts mixed up the day they saw Nicole, or the time of day.
In both these cases (and in many, many more - the Truscott case was full of such problems), witnesses are honestly reporting what they remember, but (hammering my cognitive science research here) because when they say what they saw at the time, when they didn't realize its significance and it wasn't something they paid close attention to, they got some of the details wrong, even if the memory itself was valid.
Witnesses, and the rest of us, are also prone to conflating memories, that is putting together in our recollection two things that actually happened, and that are connected, but not in precisely the way they happened. For example, I might remember vividly a conversation with a colleague and correctly recall the gist of what was discussed, but mistakenly think it occurred at a meeting we both attended instead of in the workplace. I "conflated," that is, merged together, two legitimate memories into one, which normally doesn't matter, but it can be very important if dates/times and locales are critical.
This is something pretty well everyone does, because our brains try to be efficient and put related things together for easier access, like a database. Paradoxically, recalling and relating these memories verbally tends to cement errors in our recollections, rather than remove errors; our conflated memory "overwrites" the original events.
I'm sure some lies were going around, but I expect many of what sound like lies (especially ones, as for example MM's recollection of how she heard about DM's arrest) may better be explained by our tendency to mis-remember such things and/or conflate related memories together.For example, MM may correctly remember seeing something about DM's arrest on TV, also remember having some sort of argument with MS about DM around the same time and put those two together and felt certain it happened on the Friday (IIRC) which most have agreed could not have been the case because the arrest was not on TV that night. She might, however,
remember it that way, and be telling the truth in that sense.
Not to say some testimony is not lying, but I think there are almost certainly many errors of memory that are honest. And you can be sure the witness doesn't think they are errors! We rarely notice our own errors in everyday life unless something corrects us, like checking our datebooks or calendars and realizing, oh, that didn't happen on day x, it was day y.
