Case Notes: S01E09 – The Yorkshire Ripper, part 3 of 5


See the Self-care page if you need support.

For half a decade, the Yorkshire Ripper terrorised northern England, attacking and murdering at least twenty women. In the third of this five-part miniseries, the police turn to linguists for help. Can Stanley Ellis and Jack Windsor Lewis help the police catch a serial killer? En clair is a podcast about forensic linguistics, literary detection, and language mysteries from past to present. Below you will find data, audio credits, further reading, and a transcript of the podcast.

Audio credits

Kai Engel – Paranoia
Kai Engel – Melted Wings
Lee Rosevere – Taking the Time (looped)
eddy – Building Tension

Credits, sources, and more

Letters and tapes


This is a strong content warning. All parts of this extended, multi-part episode contain strong language and/or descriptions of violent murders. Teachers, guardians, parents, and caregivers should listen to every part of this episode first, and then make their own judgement call about whether this content is suitable for younger or more sensitive audiences.

Case S01E09 – The Yorkshire Ripper, part three of five.

We resume this case in June of 1979. So far, since around 1975 or so, there have been eight or nine murders. And there have been a handful of attempted murders – possibly two or three, maybe four or five or more. The police, however, cannot be sure. Some of the attacks bear clear resemblances to others. But some seem more random, less in keeping with any overall pattern. The only consistent variable thus far is that the victims are all girls or women. Other potential links are that many of those women are, or were sex workers, but even this is inconsistent. However, because of this latter aspect, the press have dubbing the killer the Yorkshire Ripper – a throwback to Jack the Ripper, the murderer of at least five sex workers in the East End of London during the late 1800s.

The Major Incident Room at Millgarth Police Station in Leeds is being crushed under the literal weight of paperwork, and under the metaphorical weight of public and governmental pressure to make some sort of headway. And into the midst of this crucible of ever-escalating tension and frustration, have arrived three letters and one tape recording, all claiming to be from the killer himself. Initially sceptical, when a declared intention to kill an older woman in Liverpool or Manchester is then carried out, there is a sudden change. Perhaps this really is the killer? If so, this might be their first and best chance at a breakthrough. Indeed this would be the only breakthrough in almost five unrelenting years, but to capitalise on it, they need help. In their desperation, they take an almost unheard-of step, and contact a linguist called Stanley Ellis.


Welcome to en clair, an archive of forensic linguistics, literary detection, and language mysteries. You can find case notes about this episode, including credits, acknowledgements, and links to further reading at the blog. The web address is given at the end of this episode. And, if you get a moment, leave us a rating wherever you get your podcasts from.

Stanley Ellis

Stanley Ellis is a dialectologist and a phonetician, and he’s around sixty-three at the time that he becomes involved in this case. Thirty years later, in 2009, his many obituaries across national newspapers will describe him as a pioneer. He was a meticulous and dedicated researcher who, between 1962 and 1971, spent a decade on the road, first with a motorbike and sidecar, then latterly with a Land Rover towing a caravan, studying the dialects of England. This work, incidentally, remains a cornerstone publication on regional varieties. Born in Bradford in 1926, he would later be a navigator in the RAF, then a lecturer at Leeds University, the editor of a journal, a presenter on Radio 4, and an expert advisor in an ever-increasing numbers of court cases that involved voice recordings.

Ellis was arguably the first person to give expert speaker identification evidence in an English Court, all the way back in 1967, twelve years before he would find himself caught up in the frenetic whirlwind of the Yorkshire Ripper enquiry. As a result of this early involvement in forensic speech analysis, the UK’s domestic intelligence agency, MI5, recruited him as a consultant. In case it helps, and for those who like this sort of nerdy trivia, MI1 was the British Military Intelligence department within the War Office that dealt with code breaking and communications, and I’ll be doing an episode on them later. MI2 and MI3 handled geography and cartography. MI4 was aerial reconnaissance. MI5, as I’ve said, is the UK’s domestic intelligence agency, and MI6 is the UK’s foreign intelligence agency, though technically all these names are now defunct. MI6, if it helps, is the one that James Bond, 007, works for. Ellis, though, was as unlike James Bond as it is perhaps possible to imagine. Friendly, enthusiastic, and personable, perhaps the most revealing insight into his character is to be found in his delight when technology advanced such that he could conduct radio interviews in his pyjamas and slippers, from the comfort of his own study. In short, Ellis was at once a brilliant, tenacious academic and good-natured, affable human being – a combination that can be disappointingly infrequent.

Back to this case, however. About a week after the tape arrives with the police, that is, somewhere around the end of the third week in June of 1979, Ellis receives a call. The police tell him that they have a tape for him to analyse, but beyond that, they give him no background information about the circumstances leading up to the delivery of the cassette. Ellis undertakes a preliminary, impressionistic analysis of the voice on the tape, and concludes that the speaker’s dialect shows…

a predominant, if not exclusive influence from the North-East of England, but not from Tyneside or North Yorkshire. At this early, impressionistic stage in my analysis, I considered it most likely that the accent was from the Sunderland area, Sunderland being an industrial, east coast seaboard town in County Durham roughly midway between the rivers Tyne and Tees.

For those not familiar with the geography of England, North Yorkshire is a shire county that includes York, Harrogate, and Scarborough. Tyneside is a conurbation that includes Newcastle-upon-Tyne, Gateshead, and South Shields. In different words, these are both pretty big regions, and they’re both reasonably close enough to the various crime scenes to make travel between them and the locations of the victims viable. However, Ellis’ immediate analysis suggests that the person on the tapes is not a native of these locations. Meanwhile Sunderland – the area he does suspect – is considered to be the heart of a conurbation known as Wearside.

After Ellis communicates this to the police, they inform him of the existence of the three letters with their Sunderland postmarks, and that the handwriting on the parcel containing the tape has been matched to that on the letters by a document examiner. Ellis suggests that the police play the tape to locals to more precisely determine the target area. His argument is that eastern county Durham speech is excellent for fine-grained geographical variation and the locals of those areas are ideal people to judge the accent as being more or less similar to their own.

The average person is not a linguist, of course, and would never be able to describe the sample using technical and scientific jargon in the way Ellis could. But, even a lifelong expert phonetician cannot hope to have that extraordinary sensitivity to a similar accent as someone who has been born and raised into it. A layperson might not be able to say why an accent matches or does not match their own – indeed, that was Ellis’ job – but they can provide a superb linguistic barometer for saying if there is a strong resemblance to begin with.

Instead of taking the tape to Sunderland and trying to match it as closely as possible to a particular group of local speakers, the police instead decide to withdraw it. Whatever their intentions, however, the secrecy does not last for long. Within days, the tape’s existence is leaked to the press, and on the 23rd of June, 1979, details are published in the Yorkshire Evening Post – a large local broadsheet newspaper covering the region where the Yorkshire Ripper attacks are taking place. This is a publication that will come up again.

The public reaction to the news about the tape is swift, and perhaps boxed into a corner, the police hold a press conference three days later. At this press conference, the officer in charge of the Yorkshire Ripper investigation, Assistant Chief Constable George Oldfield of the West Yorkshire Police, plays the cassette tape to the general public and he appeals for help in identifying a suspect. This particular copy of the tape is an edited version, with the pauses from the original copy removed – a point I return to later. The tape is played both to the press and to the general public via the radio, and journalists soon start dubbing the speaker on the tape Wearside Jack.

Stanley Ellis chooses at this time to decline any and all interview requests – a strategy he largely maintains throughout the investigation, though one that he later questions. Unable to extract any comment from him, therefore, and desperate for another expert to give their opinion, on the very same day that the recording is released, BBC Television rush a copy of it to the home of Jack Windsor Lewis so that he can give his professional opinion on its contents.

Jack Windsor Lewis

Also born in 1926 like Stanley Ellis, Windsor Lewis is also around sixty-three at the time that the BBC breathlessly arrive on his doorstep. The likelihood is that they came to him because in 1969, some ten years earlier, he had authored the book entitled, A Guide to English Pronunciation. And, like Ellis, Windsor Lewis is a Lecturer at Leeds University. This has two advantages for the journalists – he is close at hand, which is always a bonus, and as a colleague of Stanley Ellis, he might have some inside information that they could potentially winkle out of him.

Whatever the case, Lewis gives his opinion: he describes the voice as having a Geordie accent, but he uses this term in a looser sense to include both the Tyneside and the Wearside accents. If you remember, by contrast, Ellis was clear at the outset that he does not consider the speaker to come from Tyneside. Similarly, he does not describe the accent using the term Geordie because to him, this term applies specifically to Tynesiders. The accent of Wearsiders, meanwhile, is sometimes called Makkem – another term that Ellis also does not seem to use but an interesting one to know nevertheless.

We’ll come back to Ellis’ analyses in a bit. For now, Windsor Lewis’ opinion is that this recording is both extremely unlikely to be of someone using a fake accent, put on just for the tape, because it is such a difficult accent to emulate convincingly, and moreover, he also believes that the speaker is not attempting to mask their own accent. They have simply recorded themselves speaking as they speak, and this happens to be in a remarkably unique accent which is so broad that many of the journalists attempting to transcribe the tape struggle to understand some of the words being said.

Unsurprisingly, given the astronomically high profile of the case, and this sudden release of a tape recording, Ellis and Windsor Lewis are not the only linguists contemplating the speaker’s likely origins, and out of Glasgow comes a theory, based on the slightly imperfect articulation of the word sorry, that this Wearside Jack might be a habitual stammerer. This is a theory that Windsor Lewis rejects from the outset, and it’s one that we will return to at a critical point in the story, but for now, how are the police handling this new information?

Untold consequences

Based on Ellis’ initial, preliminary analysis, the police narrow their focus to suspects who live within the North East of England, or who were raised there and then subsequently moved to West Yorkshire – the area encompassing Leeds and Bradford, where most of the attacks and murders have taken place. In the ensuing days, too, at the request of the police, Ellis appears on a radio programme with Detective Chief Inspector Dick Holland. Holland gives an account of the case to local listeners, and in turn, Ellis draws attention to key features of the speaker’s voice.

A couple of weeks later, on the 29th of July 1979, Sutcliffe is interviewed again as a result of his car being spotted in three of the red-light districts under heavy surveillance. His wife gives a loose alibi for him, and dissatisfied with the interview overall, Detective Constables Laptew and Greenwood obtain samples of his handwriting, search his home and garage, and write up a report of their suspicions which they hand in to the Major Incident Room. But here, we find some of the earliest influences of the letters and recordings on the outcome of the investigation. Sutcliffe’s writing is deemed not to match that of the letters. On this basis, he is eliminated as a suspect, and the incident room do not follow up on the report.

There is a key technical issue to point out here, which I’ve alluded to before, but now it’s worth explaining in more detail. Police can treat evidence in one of two ways. If confidence in the evidence is extremely high, they can use it to eliminate suspects. That is, they can use it to determine that there is no reasonable chance that this person is the guilty party. Here’s a grim example to illustrate this. Let’s say a tall, strong, mixed martial artist is walking down a notoriously dangerous back alley in the early hours of the morning. Their body is later found in that very alley, and the nature of the injuries is consistent with someone who has been kicked and punched to death. On this basis, one might confidently eliminate a frail, sickly, elderly individual or a child as being the attacker on the basis that they are not strong enough to carry out the crime, especially against a fit, athletic, trained fighter.

By contrast, imagine that this same person in this same alley is, instead, shot to death. One might argue that a child or frail, elderly person is unlikely to be in such a place at such a time, but unlikely is not the same as impossible, and so their age or their relative strength or where we expect them to be at three in the morning cannot be used to eliminate them in a case like this. Instead, we would use this as a line of inquiry – where were they at 3am on that night? Do they possess a gun? Did they have a grudge against this person? And so forth.

Elimination versus line of enquiry.

The very fact that the police were now using the handwriting samples – and, as we’ll go on to see later, the accent in the recording – as a means of eliminating suspects from the inquiry, indicates that at least at this point, the confidence in the authenticity of these communications was so high that the commanding officers were using it as a method of cutting down their suspect pool. As Ellis himself notes,

The view of the police was that if they caught the speaker/writer, they would in doing so have caught the murderer.

But let’s get back to the linguists and the role that they are now playing in the investigation.

A month after Sutcliffe’s latest interview with the police, Ellis is invited to Sunderland Police Headquarters to analyse the Wearside Jack tape in more detail. As mentioned before, Wearside was Ellis’ first suspicion for the origins of the accent, and he had further confirmed it by comparing the recording to speech samples he already had of school pupils from the Sunderland area, as well as some newer speech samples of police officers also from the same place. Just as a quick reminder, Sunderland is the heart of the surrounding Wearside conurbation, and Wearside is the home of the Makkem accent. Makkem is distinct from Geordie accent, which has its home in Tyneside. In case it helps you to cement the two together in your brain, Geordie is a Tyneside accent, or G&T. Gin and tonic. Makkem is a Wearside accent, or M&W. Maltesers and wine.

Mmmm. Food mnemonics.

Er, back to Ellis.

On this roadtrip to Sunderland, Ellis brings two tape recorders: one to play the Wearside Jack message, and another to record the speech of locals. Senior officers agree with Ellis that the speech on the tape resembles that of locals of Barbary Coast, a little area within Sunderland on the north bank of the River Wear. Ellis is accompanied by a young constable and he visits various local meeting places. I mean, I may be projecting here as I read his accounts of the matter but I get the sense that he is using scholarly prose to describe going to a looooot of pubs and clubs. That said, there’s an excellent rationale for this extended visitation of watering places. These are excellent places to meet lots of laypeople. British pubs and clubs are intended to be a sort of alternative living room, where you can meet up with friends and workmates, chat about whatever topic amuses you, and generally while away the time after work. They are informal, relaxed, intimate, and also open to the public, of course, so this was almost a way for Ellis to chat to people as if they were in the comfort and safety of their own houses, without actually having to go into anyone’s house. Why is this important? Because you then get a much better sample of speech from them – a more relaxed, uninhibited, faithful representation of their own accent. And in Ellis’ own words,

The response was most encouraging. Everyone we met was ready to help by listening to the tape and by commenting on it. What they said did not much matter, it was how they said it, although there were often relevant comments as to their own view of the regional origins of the voice.

Put those same people into, say, a police station, even just for an informal chat, with a gigantic tape recorder and sheets of questions on the desk and officers staring at them, and immediately the whole context is different. People then put on different identities – the one they use to keep the police at arm’s distance, for instance, or the one they use when being addressed by a scientist who is asking them all about how they speak. We’re talking here about both accommodation theory, where people choose to speak more or less like the person they’re chatting to as a method of increasing or decreasing the social distance, and observer’s paradox, where people who know they are being watched behave differently than they otherwise would. Both can and do profoundly spoil linguistic samples if what you’re after is as natural and unselfconscious a recording as possible.

In the noise and informality of a pub or club, people are more likely to get into conversation with Ellis and each other, to forget the tape recorder and their own voice, to enjoy giving their own opinions on an accent or on the murder case in general, and to provide plenty of useful data. So, whilst amusing, Ellis’ decision to hit up the pubs and clubs was actually a smart one.

Small side-story: There’s an urban legend in linguistics that one of our most prestigious researchers in the US once convinced a funding body to buy her a motorbike and a leather jacket so she could study the language of jocks and burnouts, and that she then produced a wealth of groundbreaking work from it. I’ve yet to find out if every detail of that story is true, but I hold onto the idea that this can happen to me too someday, only maybe with a new car or just some book vouchers.

Back to Ellis and his tour of Sunderland’s public houses. Ellis and his accompanying constable took a sort of L-shaped route, beginning at the mouth of the River Tyne and then initially working their way upstream, or westwards, travelling two miles from the village of Southwick to the village of Castletown. Then they turned away from the river and headed northwards to Washington, then to Hylton Castle, then to Boldon Colliery. The main goals were to get as many people to listen to the tape as possible, and then to get as many of those people talking as possible, so that samples of their speech could be recorded. So how did this work out?

Arresting articulation

Speakers in Southwick are quick to reject Wearside Jack’s accent as one of their own. Ellis notes that this could be for quite natural reasons – no one wants a serial killer to have come from their tiny part of the world, but also, the conclusion does seem to be based on accent too. By the time they reach Boldon Colliery, in a British ex-servicemen’s club they start to encounter a very different type of accent – what Ellis calls real Geordie, more like Tyneside than Wearside, and at that point, he feels they have left the target area. Later, Ellis pores over his own work on the Survey of English Dialects, and compares this with the recordings he’s just collected, plus the Wearside Jack tape. From all of this, he determines the following.

The first word in the recording, I, immediately excluded North Yorkshire and Tyneside from the list of possible areas of origins. This is because – and forgive me my phonetic sins here – the North Yorkshire I is closer to [a:], as in [a:] know. The one on the tape, however, this I is what I like to call a banana vowel, in that the sound is bendy. Real linguists would use the term, diphthong, meaning a double vowel sound. And you get triphthongs, too – triple vowel sounds, like power and royal and layer. One vowel sound blends into another, and then possibly into a third. For the non-linguists, don’t consider the spelling here. Just focus purely on the sound, with no regard for writing. So, think of the word eye, as in, your eyeball. It goes from an [a] to an [i] sound. E-y-e. It bends. Or, you know, banana vowel. So compare [a: no:] with [ai nou]. In some accents, some words have monophthongs – a single, flat vowel, like no, and in other accents, that very same word will have a diphthong, or a bendy vowel, like no. No – monophthong. No – dipththong. And if you have kids kicking about, you might even hear that rarest of beasts in the wild, the one-syllable word that is tortured into a whole triphthong, no-uh! Go tidy your room. No-uh! Don’t you dare triphthong me!

Anyway, Ellis now firmly discounts Tyneside and North Yorkshire based on other similar artefacts, including the pronunciation of the word strike. Similarly, Sunderland is counted in on the basis of the sound in words like house, down, and about – that’s an ow sound in my accent, but it can be realised as an oo (aboot) or an oa (aboat) sound in other accents. This said, Ellis does not count all of Sunderland. In his comparisons of individual words and sounds, he trawls over not just contemporary dialect and accent surveys and studies, but also across hundreds of years of linguistic history, from Old English to present. In doing so, he looks at how shifts in pronunciation have softened, hardened, rounded, flattened, shortened, lengthened, or otherwise shaped the various sound components in language. Each potential point is compared with his recordings, and from this, he is gradually able to draw lines on a map that constitute boundaries where these various features no longer occur. In this way, he is able to create a linguistic perimeter around a very small area within Sunderland. Overall, at the end of these endeavours, Ellis notes the following:

We moved further inland to the small village of Castletown and in a pub there we met a retired man whose segmental pronunciations, intonation patterns, rhythm and tempo closely resembled those of the questioned speaker.

By contrast, the further north of Castletown they went, the fewer similarities they found.

Profiling a predator

Ellis and his accompanying constable return to Sunderland Headquarters in the late afternoon, where an army of journalists are waiting for any small nugget of information that Ellis is willing to provide. Ellis declines to give any interviews, but the Wearside reporters already knew the likely area and so this fact was widely reported in spite of his reticence.

Elis summarises his report to the police that day as follows:

I stressed that although the speaker might not still live in the Sunderland area, he may nevertheless have retained contacts with that area, a theory supported by the postmarks which I had not initially known about. A local publicity campaign was initiated and the broadcast of the tape on the local radio stations resulted in a host of reports of the voice having been heard. This led to a great swamping of police resources. In summary, I reported to the police that in my opinion the man’s voice represented someone who had been brought up in the Southwick or Castletown areas, but that I had reservations concerning the possibility of his no longer living there.

And as I’ve mentioned before, Ellis stresses that the accent is not real Geordie, because Geordie is Tyneside, and this speaker does not seem to be from Tyneside. A few days after this, Ellis goes abroad, hoping that, given the uniqueness of the voice, by the time he returns from his holiday the case will have been solved. Whilst abroad, the case does indeed return to the front pages, but not because the murderer has been caught.

Witchcraft analogy

On Sunday the 02nd of September, 1979, whilst Ellis is still abroad, Sutcliffe kills again. This time his victim is 20-year-old university student, Barbara Leach. He strikes her in the head with a hammer once, and then stabs her eight times. Though not committed in a red-light area, the murder is immediately linked to the others in the series, but the Millgarth incident room, already understaffed and sinking under a mountainous backlog of enquiries, starts to buckle.

Meanwhile, in conversation with a detective, Windsor Lewis discovers that nobody has yet conducted any sort of linguistic analysis on the content of the three letters. He offers to do so, and is shown the letters on two occasions. Once, they are brought to his home in the evening, and he is allowed two hours to study them, before they are withdrawn. A second time, he is allowed sequestered access to them at Leeds Police Headquarters for a full afternoon. Windsor Lewis writes up a 1,400 words report, and in it he gives his impressions of the three letters, which he describes in his own words as follows:

The envelope in which the tape had been sent was agreed by handwriting experts to have been addressed by the same writer and there was obviously complete harmony in both content and linguistic features between the tape and the letters. Like the letters, it harped on certain localities, such as the Leeds suburb of Chapeltown, Bradford, and Manchester. It had the same simple-minded taunting and boastful manner and the same unsophisticated word choices including fairly faded types of slang such as copper, knocking about, nicked and top myself.

This writer … was thoroughly functionally literate, in the sense that he could perfectly well express himself in writing, but rather illiterate in terms of his ability, or perhaps one should say inclination, to conform to the accepted standards of written style. The letters were all written on sheets taken from pads of lined paper, no doubt with a ball-point pen. The first two occupied two pages each. There were no margins and no paragraphs, though each contained a postscript, designated as such in only one case. The handwriting was adequately legible for the most part but frequently clumsily executed. The punctuation was well below the standard that one would hope Ordinary Level examination candidates would achieve. It was often difficult to judge whether a particular mark on the paper was meant as a full stop or a comma (or whether it had any significance at all) but there certainly weren’t enough of either by the ordinary canons. About a third of the sentences had no full stops. About a third began without a capital letter, but initial capitals regularly appeared on names (though not internally in McDonald), and the pronoun ‘I’ was always capitalized. Of the three opening salutations, two had full stops rather than the conventional comma, and the third was unpunctuated. The placing of apostrophes, where they were bothered with at all, was highly erratic. No use was made of exclamation or question marks (the latter orthodoxly requisite in only two places) or of parentheses, though there was one use of what is in fact apparently (among schoolchildren at least) quite a common punctuational folk usage whereby inverted commas (quotation marks) are employed as parentheses. These occurred around a man I respect at the beginning of letter no. 2. At the one place where quotation marks were requisite they were not employed.

What was very notably idiosyncratic was the frequent tendency to leave a space before completing a contracted spelling with its final nt, etc. The spelling appeared to suggest perhaps rather a contemptuous attitude to orthodox orthography than just incompetence. The forms nite and cause (used five times and with no use of the ordinary spelling because) are obviously simply defiance of orthodox usage. This was perhaps the case with I not for Im not (twice) and even possibly they’re learn for they’ll learn and to hot for too hot (which last was spelt correctly on another occasion). All of these last three phrases could be read back as perfectly normal spoken forms in the English of a speaker of the sort heard on the tape. Indeed, only the weakening of the adverb too to a schwa vowel (in the way general English treats the preposition to) is confined to northern English dialect among these usages. The extra s in obvisous, the superfluous h in my one regret his, the missing r from Huddesfield, the missing final s from them whore, the missing word is from between My purpose and to rid and the failure to use an initial capital for Ripper at one point were all no doubt merely slips.

In general the literary style was an unremarkable, clumsy mixture of colloquialisms, dialectalisms, slang, telegraphese and journalese, but it did contain some strikingly idiosyncratic items. The most extraordinary of these appeared to be the adjective cursen. This is so completely unattested a form that one would unhesitatingly have dismissed it as a slip of the pen for cursed were it not for the fact that that very word was used in the immediately previous sentence, and there spelled curserred Such a spelling is a quite unsurprising ‘illiteracy’ for cursed, whose unstressed vowel could be expected to have the schwa (mid-central) quality that such a spelling suggests in the area in question, where, for example, patted and pattered are not necessarily distinct. (Similarly the orthographical form cursen could as well represent the grammatical form cursing.) If it was a genuine usage it was certainly a very rare and distinctive one. There was little else that was particularly distinctive, though the slang expression close call (like top myself) is by no means equally well known in all areas (unlike e.g. daft, copper, kidding or collared). The expression lassie(s), which occurred several times, is one which is also far from universal in northeast England. Indeed, it is likely to strike most English people as mainly a Scottish usage. Incidentally, lasses would be phonetically quite distinct from lassies in this area.

Some of the grammatical usages suggest the bottom of the social scale, e.g. them sluts, them whore(s), should have took, others simply the stumbling of an untrained writer as at Its yours and theirs fault. The verbal style was reminiscent in one or two places of a letter attributed to the original London Jack the Ripper in which murders are referred to as work and the expressions [joke] gives me fits and [I gave] the lady [no time to squeal] occur. The most strikingly distinctive trait in the whole of the letters was, however, not so much linguistic as graphological. They contained four occurrences of the numeral 7. As it happened, the only one of these which appeared in the published extracts was in its normal British form, but in the three other occurrences the numeral was formed in the very alien continental way with a horizontal crossbar half-way down. This could have suggested residence abroad, but it could also have been acquired at school. In view of that possibility, it seemed to be very well worthwhile communicating to the public the locality, in the hope that it might give some sort of lead in the direction of the person who had written the letters and recorded the tape and whom the direction of the police investigation into the case insisted they were convinced was the murderer.

When Ellis returns from his holiday, he soon discovers that police are interviewing men with Geordie accents about the attacks and murders, and that they are, moreover, eliminating them on the basis of the handwriting in the letters, and the accent on the tape. Dismayed, Ellis calls Windsor Lewis. Both linguists agree that the speaker’s voice on the tape is extremely distinctive and should have been identified now by a family member, a friend, or a colleague. More substantively, however, they agree that the tape is likely a hoax and not connected with the crimes. In other words, in their view, Wearside Jack and the Yorkshire Ripper are almost certainly two entirely different people.

On the 13th of September, eleven days after the last murder, West Yorkshire Metropolitan Police issue a Special Notice to all the forces in the country giving details of the sixteen murders and attempted murders thought to be linked to the Yorkshire Ripper series. These include details of the letters and tape, but exclude the descriptions and photofits of the suspect. All the forces are told they can eliminate suspects – and remember here what I said about elimination versus further lines of enquiry – that they can eliminate suspects if they (are):

A) born before 1924 or after 1959;
B) obviously not Caucasian;
C) wear a shoe size 9 or above;
D) have a blood group other than B, or;
E) do not have a North East (Geordie) accent

Though this elimination list is kept secret, Stanley Ellis happens to see a printed copy of it, but when the Chief Constable is asked about it, he denies its existence on multiple occasions.

By coincidence, the very next day, on the 14th of September, the incident room in Sunderland receives a call. The anonymous person asks for a message to be passed on to Oldfield that the letters and taps are fake. Constable Keith Mount, who answers the phone, is convinced that this is Wearside Jack, and even manages to record part of the conversation on a hand-held device. That recording is then supposedly analysed by a Home Office lab, but they rule that the caller’s voice does not match Wearside Jack’s. Whether this is true or not, the ultimate fate of that recording is unknown for several years until, remarkably, it arrives at the forensic acoustic and speech science lab of J.P. French Associates, who provide a transcript of the part of the call that the officer managed to record. That transcript runs thus:

PO           Repeat that, I can’t hear you, it’s a bad line.

C              Tell him it’s a fake.

PO           What’s a fake?

C              The tape recording.

PO           What one is this? The one that he’s just received?

C              The Ripper tape recording.

PO           Aha. How do you know that?

C              Just tell him.

PO           Just tell him?

C              The one in June.

PO           Pardon?

C              The one in June.

PO           I’m sorry it’s a bad

C              … sorry.

PO           It’s a bad line, you’re going to have to repeat it.

[Caller hangs up]

In the middle of September, both Ellis and Windsor Lewis submit their reports to the police. In Ellis’s report, he stresses that the handwriting and voice investigations should be dealt with separately from the murder enquiry, as he cannot be sure they are linked. In short, he strongly intimates that the letters and call are a hoax. However, externally, there is no sign that they have been attended to, Windsor Lewis states that his report is simply sat on:

Although I gave various reminders of how useful I hoped it might be, I merely eventually got a telephone call from a senior detective on the case thanking me for the trouble I had gone to in providing it.

In short, whilst the expertise of both linguists seems to be used as a method of elimination, their concerns that this is a fatal error and that the letters and tape are a hoax appear to be entirely disregarded.

End of part three of five.


This episode of en clair was researched and fact-checked by Rebecca Jagodziński. And it was scripted, narrated, and produced by me, Dr Claire Hardaker. However this work wouldn’t exist in its current form without the prior effort of many others. You can find acknowledgements and references for those people at the blog. Also there you can find data, links, articles, pictures, older cases, and more besides.

The address for the blog is And you can follow the podcast on Twitter at _enclair. If you like, you can follow Rebecca on Twitter at RjJagodzinski, and you can follow me on Twitter at DrClaireH.