An Opinion on the Timing Problem
by Martin Shough
The timing of these events has always been a matter of debate due to what appears to be a material discrepancy in the 'classical' sources. Teletype BOI-485 gives "140010Z through 140330Z" for the total duration of events at Lakenheath, whereas Perkins maintained that these events began during his 131700Z-140000Z shift. Why should a disagreement about the time be so contentious? In particular, why should Perkins have remained so stubborn in the face of this documentary evidence?
Perkins had freely admitted from the start that he couldn't be sure of the exact date, for example. When the official report was found he accepted the date without demurral. There was no controversy, because he accepted that his recollection of the date was vague.What did he have to gain now by contradicting that same official report on this apparently trivial question of the time? On the face of it he had nothing to gain by courting controversy - unless he was very sure. If he was very sure then he did have something to lose by just deferring to BOI-485 - integrity.
In August 1956 the intelligence officer who prepared BOI-485, Capt. Paula Stimson, himself drew cryptic attention to an internal timing discrepancy in the Lakenheath evidence sources, with the remark:
Thus two radar sets and three ground observers report substantially same. Radars reported these facts to occur at later hours than the ground observers.
In November a confused Blue Book officer remarked on the case file record card that this statement (about Lakenheath) appeared to be corroborated by the absence of simultaneous radar-visual reports in the document IR-1-56 (about Bentwaters) which had belatedly - and somewhat mysteriously (see below) - just been received from Bentwaters via Washington. More judiciously, Blue Book consultant Dr.Hynek noted in an internal 'evaluation' memo that the statement contradicted explicit reports of radar-visual simultaneity elsewhere in BOI-485 and so needed clarification.
In 1968 G. David Thayer mused anew on the possible meaning of Capt. Stimson's remark in the Condon Report [Thayer 1969]. By this time it seemed curious, but academic, inasmuch as the 'classic' Bentwaters/Lakenheath scenario that now emerged, based largely on Perkins narrative with IR-1-56 and BOI-485 adduced in ad hoc support, grew in status with the retelling. But during the 1970s critics such as Philip Klass pointed out that Perkins' independent account also imported a time discrepancy, because it contradicted the start time recorded in BOI-485. This created a serious structural problem for the scenario carefully developed by Thayer, McDonald [McDonald 1972] and others. The implied delay between the first telephone alert and the 0010 start of the action at Lakenheath - possibly as much as two hours on Klass's interpretation - didn't add up [Klass 1974].
In 1975 Perkins insisted in a letter to Stanton Friedman that it didn't add up only because BOI-485 was wrong:
We had been working the UFO and the first intercept before shift change which would establish the telegraphic report as being in error . . . .
In debate with Gordon David Thayer and Stanton Friedman in 1976, Klass pursued this as an issue that spoke to the credibility of Perkins' recall. He suggested that Perkins must have misremembered his shift time. But Perkins in a letter to Thayer continued to maintain that
I was working from 5PM to Midnight . . . The time in the telegraphic report is wrong if it does not agree with me. I received . . . the call around 1000 or 1100PM.
In his first letter to Friedman Perkins had insisted that there was only a "small delay . . . ten to twenty minutes, certainly not more" between the telephone alert and the appearance of the target, and he listed names and full service numbers of personnel including three from the midnight shift who he said "made relief in the middle of all this" and who therefore might have been "in on the last part". (One of these, Sgt. Thomas Emerick, was repeatedly contacted, by this author and others, but declined to become involved and is now deceased. Another, Sgt. John Foundos, has not been traced but was certainly on the RATCC staff and is mentioned in an August 1956 Lakenheath USAF base newspaper article in the author's possession. Two others, Richard Lynch and John Bell, were traced by the author with the aid of a US military officer locator service but have not responded to letters.) According to Perkins the relief time for the midnight shift was 2345 with a 15-minute handover and he stated: "we had been working with the problem for an hour or two before this occurred." In a letter to the present author in 1987 he certainly had not budged from this position:
I was working from 5PM to midnite - the evening shift. We were relieved by the Midnight Shift (1200-0700AM) at midnite - they arrived at 1145PM but we stayed to help out till at least 1230AM, then after they took over completely we still stayed to make out the report in what had happened, consolidate the maps we had made of the UFO and Intercept - so it was probably 1:30 or 2:00AM before we left the RATCC.
In this matter Perkins was nothing if not consistent.
Now clearly if Perkins had not been on the 1700-midnight shift as he believed, but was instead on the midnight-0700 shift, then his recollection of a 10-20 minute delay between the moment when he had his shift 'start scanning the scopes' and the acquisition of the stationary target could be squared nicely with the 140010 start time logged in BOI-485. Could he have been wrong? Could it be, as Klass argued, that Perkins' memory was confused because of a floating shift arrangement? A fixed shift pattern would be clearly memorable; one which rotated through the week might not be. Moreover, if Perkins took over at midnight then this would mean that the 2255 telephone alert would have arrived during the previous shift, and this could also explain Perkins' ostensible "error" in believing that it came from Sculthorpe rather than from Bentwaters.
Thayer responded that if this were correct then the next shift could not possibly have been involved at all, still less have "made relief in the middle" as previously stated by Perkins, since they would arrive at 0645 for an 0700 start and according to BOI-485 itself the whole affair was over by 0330. The matter of a shift making relief in the middle of the action is a strong structural feature of Perkins testimony, and one in my opinion unlikely to have been misremembered. If this were the only strong feature of that testimony it would in itself be sufficient to demonstrate that at least one of the two times given in BOI-485 must be wrong.
This reasoning suggests that BOI-485 probably cannot be relied upon to contradict what the principal witness stoically maintained for over twenty years, and that the start time of 140010 is probably wrong. Now the idea that an official report to Blue Book might contain an error is hardly a revolutionary suggestion, of course. But for it to be valid to discount a detail in a documentary source one would like to have some hypothesis. The suggestion that BOI-485 is wrong is ad hoc at this stage. Why should this start time be wrong? It is clearly not simply a typographical slip, being given twice, each time with the interlocked date, i.e. 14 [th August] 0010. Evidently this date-time group was accurately transcribed from some source report or other. Therefore we are driven to ask: If this source contained an inappropriate time, was the source itself in error? Or was it an inappropriate source for the purposes of BOI-485?
This last question implies that separate reports may have been generated, possibly even by different activities of the USAF establishment at Lakenheath, appropriate to different time frames. Is this plausible? Certainly one of the lessons of recent diligent field work by members of the Lakenheath Collaboration (in particular Dave Clarke and Andy Roberts, although I should emphasise that they each have their own different views on the meaning of this 'timing problem') is that what is usually thought of as 'the Lakenheath incident' is just the most visible part(s) of a rather large and rather vague iceberg of previously unknown events taking place at various military airfields around that time. That multiple independent incidents may have happened in this area of East Anglia on the night of August 13-14 is not only possible but is in fact now supported by documentary evidence.
Those matters are discussed in their place. Returning to the debate as of 1976, if we look into the future only as far as 1978 we know that the RAF Fighter Controller's independent account of Venoms intercepting a Lakenheath UFO will then be found to interlock well with Perkins' as to time, stating that his facility at Neatishead was alerted between 30 and 60 minutes after the start of his shift, which began at 2300Z, August 13. Evidently, the radar sightings which led to this alert could not have occurred after midnight and so this evidence conflicts with the timing in BOI-485.
But then again, looking even further into the future (1995) we find that the log books of two 23 Squadron Venom aircrews who attempted to intercept a Lakenheath UFO that night will turn out to clearly record the time of the first scramble as 0200. Evidently this time is consistent with the times given in BOI-485, but is definitely inconsistent with Perkins' testimony and the implications of the RATCC shift pattern.
What is the relationship of these facts? If the times associated with these events were the only point at issue it might be tempting to try to force fit them. Indeed, in later testimony the RAF Fighter Controller grudgingly began to improvise a later shift time after being faced with the apparently incontrovertible evidence of the 23 Squadron log books. But this was not his genuine memory, and he need not have felt this pressure. Because the significant point to note about these two above sources - one from Venom aircrews of 23 Squadron Waterbeach who were on alert that night, and the other from the Neatishead GCI radar facility tasked with controlling interceptions - do not describe the same event.
This issue is discussed in detail elsewhere; suffice to say here that despite face-value appearances and a priori likelihoods it can be definitely stated that the two 23 Squadron Venoms scrambled after a UFO at 140200 were not under Neatishead GCI control, and therefore were not the aircraft scrambled by Sector and vectored by Neatishead GCI to intercept the Lakenheath UFO described by the RAF Fighter Controller. Interpretation of this fact is obviously not simple. Nevertheless, this is one of the few structural facts that can be considered clearly established, and, whatever interpretation we favour, it is in my opinion fundamental to understanding what happened that night.
The above facts argue that it might be wise to begin with a presumption in favour of the possibility that events around Lakenheath that night fell into two, or more, distinct phases, distinguished perhaps by different personnel and/or facilities, perhaps with different reporting protocols and in different time frames. With this presumption certain conflicts begin to make sense.
Perkins recalled that his retiring shift was starting to make out its report on the core incident (the interception) by about 0030 when things were winding down, and then left the building after another hour or so. It is evident that this report could not relate to events which occurred over the next couple of hours and ended when "all targets disappeared from scopes at approximately 0330Z." (BOI-485) Therefore the reference to 0330 (given separately four times in BOI-485 for the end of the incident) indicates that the compiler drew this information from other sources. Presumably these sources would include a report prepared by the RATCC relief shift describing targets observed by them after the hand-over.
BOI-485 states that it is based on records of "personnel interviewed and logs of RATCC" and includes evidence both from RATCC radar and from Lakenheath GCA radar. Evidently the "personnel interviewed" included GCA radar operators. Perkins was not interviewed. His detailed written report went to the officer in charge of the RATCC (not Capt. Stimson but his 1979th AACS Communications Squadron Commander) who undertook to "take care of it". After that Perkins "heard no more about it." This report, with accompanying radarscope maps and tape-recordings made continuously throughout the incident, should have been forwarded to Blue Book in accordance with AFR 200-2, but all this material disappeared somewhere in the reporting chain and may even never have reached Captain Stimson's staff.
So it is possible that the entry in BOI-485 against question 2.f, "Length of Time in Sight", reading "Objects were observed intermittently by RAF Station Lakenheath radars from 140010Z to 140330Z", originates in the report of the relief shift Supervisor, these times being copied across into BOI-485 (or another source document thereof) and simply repeated. Note that the radar tracking reported by Perkins during his shift involved a single radar target, which was not in any way "intermittent", and that 0010 fits quite well with Perkins' recollection of the period during which hand-over was being progressively made. These two separate time-frames might also explain the inclusion in BOI-485 of at least one RATCC tracking, and others vaguely described which may relate to either RATCC or GCA radars or both, that appear to have nothing to do with the core incident described by Perkins (or the RAF Fighter Controller).
But this is not the end of the matter. Recent investigation by Dave Clarke and Andy Roberts has uncovered the 23 Squadron Diary entry for the night in question, which states that the two Venoms scrambled after 0200 at the behest of the Americans were controlled by Lakenheath GCA radar. The Ground Controlled Approach unit would be a radar trailer and standby van out on the airfield, quite distinct in location, function and staffing from the Air Traffic Control unit which was located in a room in the Tower building.
Mostly the RAF airmen themselves have no strong recollection of which facility they were in communication with, other than the fact that it was an American radar unit at Lakenheath. When they were scrambled they were supplied with a frequency (one of a number of pre-set VHF channels) and a call-sign, but nothing more. Only one of them, Ivan Logan, volunteered an opinion. In an interview with Dave Clarke he three times, unprompted, identified the controlling radar as "Lakenheath Approach radar".
It is quite possible that Logan, and the writer of the Squadron Diary entry, were both casual about the distinction between Lakenheath's airfield landing radar and its long-range Air Traffic Control radar. Nevertheless the fact remains that the only RAF documentary source referring to the American radar identifies it as Lakenheath GCA.
On its own this is a small point and carries little weight. However it is consistent with the arguments set out above. If there were in fact two 'Lakenheath incidents' that night, so that the 0200 scrambles generated a different report - not only from a different shift but from a different and physically remote facility - then, when BOI-485 was assembled two days later from what may have been a melange of partly-digested information relating to a confused series of events, the opportunity may have existed for conflation of two different time-frames.
Another point of view which has been considered in the light of the new evidence is that the date-time groups given in BOI-485 as GMT are actually based on British Summer Time (this was originally suggested some years ago, by Brad Sparks I believe). If BOI-485 is consistently in error by one hour then "140010" becomes 2310Z on the 13th, within Perkins' evening shift, and moreover, with "2255Z" being really BST for 2155Z, then (allowing for a few minutes) this could explain the puzzle of why IR-1-56 from Bentwaters times its last radar UFO at "about 2200Z", not 2255Z as reported in BOI-485.
There are some ostensible difficulties. Firstly, there is no sensible reason why a systemmatic error of this kind should run through a report from a major military operation which would be locked into the USAFE standard practice of record-keeping by GMT. Secondly, this correction would only shift the problem of the hour-and-a-quarter hiatus in BOI-485 between the Bentwaters alert and the first Lakenheath radar track (Perkins insisted that only 10 or "certainly no more" than 20 minutes elapsed). And thirdly, if BOI-485 is systemmatically an hour out, then an end time of 0330 for radar events at Lakenheath becomes 0230Z; but we know from the 23 Squadron RAF pilots' log books that the last interception of the evening under Lakenheath control had yet to begin at that time.
Now of course it is an implication of the new evidence offered here that the 23 Squadron interception referred to, handled by GCA radar, may have had nothing to do with the controllers in the RATCC and may not have been known to them. Given this, then we are saying that the timing of these later events becomes immaterial, if the time frame of BOI-485 is after all derived from the "logs of RATCC" referred to by Capt. Stimson, and if the times therein are erroneously given in BST. So this construction would favour Brad Sparks' opinion that the "2255 [BST]" Bentwaters event reported in BOI-485 is merely a corrupted version of the "about 2200Z" (= 2155Z) fast radar track described in IR-1-56. This simplification would be achieved at the cost of
There might be a middle way that captures plausible features of these points of view, removes the need to assume a systemmatic imposition of BST in the intelligence report, removes the conflict with Perkins' testimony, and preserves the Bentwaters radar-visual as reported in both sources.
The discrepancy between Perkins and BOI-485 is very close to one hour. A start time of 14:0010 [BST] or 13:2310 [Z] would sit very well with Perkins' statement that RATCC radar tracking began "10 or 20 minutes, certainly no more" after the Bentwaters alert. But this is not consistent with a systemmatic error throughout the report and requires that only the group "140010" is subject to a one hour error. Is there some reason for a confusion between military time and local civil time which would selectively target this one date-time group but not others? .
To USAF personnel in the radar room all times will be unambiguous because they are routinely working with other military personnel who are also using GMT and everyone knows that, say, 2200 means just 2200, military time, year-round. The fact that people in the civil world start calling it 2300 during the British summer would be immaterial. Civil time would float against fixed military time by and large without being noticed. However, the fact that civil time would float against a fixed shift pattern (the opposite of what occurs in the RAF, where shift patterns float with BST to follow civil time) might be noticed. The reason is that the shift changes at 'midnight', but there are two 'midnights' - there is the midnight of GMT, 0000Z military time, and the civil midnight, which will be 2300Z. If a serviceman says "2359" you wouldn't need to ask "which one?"; but if he says "midnight" you might.
So it happens that of the various times given in BOI-485 the start time "14:0010", or ten-past-midnight, uniquely has this potential for ambiguity. Is it a coincidence that this time is the cause of all the trouble? Is it a coincidence that correcting for a hypothetical one hour error here produces the time - 2310Z - that fits the context so well?
The answer has to be another question: Why would any USAF serviceman recording or reporting this time translate it as "ten past [civil] midnight" in the first place? This only seems plausible if it coincided with a nominal "midnight shift" change in a shift pattern wedded to local time. Could this be the case? If so, then we are back to the problems of the systemmatic error theory.
Both versions of the BST hypothesis are therefore at best inconclusive. Nevertheless it remains true that the timing is a fly in the ointment from every point of view. I accept that the scenario suggested here is speculative - and is not without problems despite other persuasive arguments in favour of a 'missing interception' - but some interpretation which honestly resolves these issues seems to be necessary. And developments in 2003 concerning the hitherto unknown involvement that night of at least one other aircraft, piloted by the 23 Squadron senior officer, Wing Commander A. N. Davis, suggest that this speculation is on the right track.
© Martin Shough 2003