Subject: Formal complaint — GGAT / Heneb: Great House Farm (Tŷ Mawr), Llandough — HER identifiers and the 1978–1995 record
From: Sion Buckler
Direction: sent
Date: Sun, 27 Sep 2026 04:15 UTC
Institution: Heneb (formerly GGAT) / Historic Environment Record
Refs: none
Archive id: gm-1a0e11391052951a
------------------------------------------------------------

Chief Executive / Complaints Officer,

This is a formal complaint to Heneb: The Trust for Welsh Archaeology, as
operational successor to the Glamorgan-Gwent Archaeological Trust, and to
Glamorgan-Gwent Archaeological Trust Limited (The), company 01276976, which
remains an active company with its registered office at Heneb's address.

It concerns Great House Farm / Tŷ Mawr, Llandough, Vale of Glamorgan (near
Penarth / Cardiff); ST 1681 7331; NPRN 20182. Cadw and RCAHMW have supplied
three unreconciled HER references for the same site: GGAT02038s, GGAT03586s
and GGATE003872.

This is not a chaser of my records request of 8 September 2026 or the
schedule of 19 September (ref GHF/HENEB/2026-09-20). That request remains
open and is due a response by 6 October. Please treat this complaint as a
separate file and give it a complaint reference.

What this is not. It is not an allegation that GGAT demolished the house,
advised that it be demolished, or decided that it should not be listed. It
is not an allegation of dishonesty, bad faith or criminal conduct. It is
not a title claim and I do not ask you to adjudicate BP v Buckler. GGAT's
letters of 13 and 15 December 1993, recommending a Breach of Condition
Notice when Condition 18 had not been met, were the Trust doing its
curatorial job. That is recorded here so it is not lost in what follows.

The charge, put narrowly. GGAT was the public curator of the regional
record; the adviser who told the planning authority it could not determine
89/01396/OUT without an archaeological assessment; the body which, in the
same document, specified machine-trench evaluation, quoted £2,000 and two
weeks, and recorded that it was already negotiating that work with BP
Properties; the contractor who then executed that evaluation on a
churchside monastic site its own 1978–80 excavations and its own 1982
letter had already flagged; and the author of a zoning map which treated as
empty ground in which, four years later, inhumations were found a few
centimetres below the topsoil. Disclosure of the dual role was not
management of it. The public record still does not reliably carry what was
found. Heneb's entire substantive reply since May is one sentence about
corporate form.

I enclose GGAT — Forensic Critique, Version 4.0.0 (21 September 2026). It
distinguishes established fact, reasonable inference, what cannot be
established, and facts that cut against the criticism. Where earlier
versions were wrong, the corrections are printed and kept. The critique is
the sourcing. This letter is the complaint. Please answer the letter. Treat
the critique as the schedule of sources, not as a substitute for numbered
answers.

1. I have not seen the HER

I do not assert the wording of GGAT02038s, because I have not been allowed
to read it. Cadw has said the content of the HER is managed by Heneb.
RCAHMW and Cadw have given three different identifiers for the same site.
Tests of Archwilio on 20 September 2026 did not return an error for
unresolved records. They returned a different record, from a different
trust, for a different site, with no warning:

GGAT02038s → GAT46242 (Gwynedd desk-based assessment)
GGAT03586s → GAT92911 (farmstead, Warwick Place, Gwynedd)
GGAT02038s, alternative path, three attempts → E001657 each time

A control query for an indexed record returned correctly. Please test those
three queries, correct the substitution if confirmed, and tell me the
outcome. Then permit inspection of the source file and supply an index of
every linked site and event record with its correct reference.

2. "Partially recorded" is RCAHMW's phrase, not a finding I have verified
on the HER

On 10 June 2026 Dr Adam Coward quoted NMR 6048945: "The house remained
partially recorded because of problems created by an ownership dispute. The
house and farm buildings were suddenly and completely destroyed by B. P.
Properties Ltd. on 6th December 1988 amid considerable local controversy."

That is the Royal Commission's file. I hold H. J. Thomas's measured
ground-floor plan of 26 April 1974, at 1:100, identifying the principal
rooms including the Dairy and recording the beams. GGAT reproduced that
plan as Figure 2 of its 1990 assessment, by kind permission. A measured
ground-floor survey that could not reach the first floor and roof is not,
without more, a failed survey.

Please state, from records you hold:
(a) whether the HER uses the phrase "partially recorded", and if so from
which source;
(b) what parts of the building the 1974 survey measured, and what it did
not;
(c) what contemporaneous document, if any beyond NMR 6048945, establishes
that access was restricted by the ownership dispute, by whom, on what date,
and to whom;
(d) whether Trench 7 of 1990 (north-east corner / former dairy, stone
floor, mortared partition) was ever reconciled against the 1974 plan, and
if the Trench 7 drawings, photographs and grid survive.

3. What GGAT held in 1988, and what it did not do

On 1 October 1982 GGAT told the Vale that occupation might continue north
of the church and that thorough investigative work should precede
development. Between that letter and November 1988 no GGAT act concerning
this site has been identified. RCAHMW added to its 1974 plan on 12 December
1988, six days after the demolition. There is no equivalent GGAT event. The
1990 report notes architectural features "observed in the rubble" and then
begins GGAT's own account with the words "Following the demolition… GGAT
entered into discussion with Debenham Tewson".

Please state what the SMR entry contained immediately before December 1988;
whether the 1982 advice was on it; whether any emergency building record,
watching brief on the demolition or the rubble, or post-demolition fabric
report exists; and where the carved stone and other architectural material
later referred to now is.

4. "County treasure"

GGAT 1989 §2.3: "At present our information on the Great House Farm area is
limited to the records of the recently demolished property which was listed
as a 'county treasure'." The phrase conferred no statutory protection.
Identify the list or card, the date, the criteria, and what information
about the building accompanied it. If you do not hold that document, say so.

5. Specify, price, bid, execute

In one 1989 consultee report GGAT advised that 89/01396/OUT should not be
determined without an assessment; specified machine-excavator trial
trenching; quoted two weeks at £2,000; and recorded that negotiations with
BP Properties via Debenham Tewson were already under way. BP then
commissioned GGAT. The 1990 report disclosed a "dual role" and invited
either party to go elsewhere. Disclosure is recorded. It is not the same as
management.

The IFA Code of approved practice for contractual arrangements was adopted
on 17 September 1990, after the assessment was delivered. It does not bind
that commission. The Welsh Archaeological Trusts' Code of Practice dates
from November 1997 and is not applied retrospectively. What did bind the
1989–90 arrangement, as contemporaneous standards rather than as a pleaded
cause of action: the IFA Code of Conduct in force from June 1985; the BADLG
Code, which GGAT expressly endorsed in the 1989 report; section 13 of the
Supply of Goods and Services Act 1982 as the implied standard of care
between GGAT and BP (the family was not a party); and the fiduciary duty of
charity trustees to identify and manage conflicts, not merely to notify two
commercial counterparties.

On 18 May 1994 GGAT's Curatorial Section wrote, of another contractor: "An
archaeological contractor should never be in the position of writing his
own Brief as well as carrying out the work." That principle was not met by
GGAT's own 1989–90 arrangement.

Please identify who wrote the 1989 recommendation, who accepted the
commission, whether those functions were separated, whether trustees or
Cadw were told, and who wrote the 1990 brief. If you hold Appendix "B" to
the officer's report (GGAT's further comments of 8 February 1990), Appendix
"D" (the representation from a County Councillor identified as a member of
GGAT), or Appendix "E" (the applicant's agents' written response to GGAT's
report), produce them. None of those three is in the planning file released
in 2026. Appendix "A" is. I do not assert that GGAT removed them. I ask
whether copies survive in GGAT/Heneb files.

6. The 1990 method against GGAT's own model

GGAT had excavated an early-medieval cemetery, a Roman villa and a
Tewkesbury Abbey grange immediately south of the same church, on the same
alkaline Liassic geology, and had written that occupation might continue
north of it. It chose eight machine-cut trenches, some placed "at random",
and then rated uninvestigated eastern ground "non existent to low"
potential. That rating converted a logistical limit into an empirical
claim. Cotswold's December 1994 preliminary report recorded that burials
continued into that zone and "lay only a few centimetres below the topsoil".

I do not say a 1990 evaluation had to predict 1,026 burials. I ask whether
the method was capable of detecting shallow inhumations on ground where
GGAT's own experience predicted excellent bone survival, and whether the
1990 evaluation has ever been reviewed against the 1994 results. If it has
not, will Heneb review it now.

7. Human remains and the 1981 Act

GGAT identified the Disused Burial Grounds (Amendment) Act 1981 and found
disarticulated human bone by the churchyard wall. That Act turns on whether
the land is a burial ground. A colour-coded potential map cannot answer it.
What legal advice was obtained, when, and from whom? What evidence set the
burial-ground boundary? Was a review undertaken after 1994?

8. Two burial figures, both correct

Cotswold's December 1994 preliminary report, §4.2: 858 inhumations and 152
disarticulated bone groups. Holbrook and Thomas, Medieval Archaeology 49
(2005): 1,026 burials comprising 814 articulated inhumations and 212
disarticulated bone groups, of which 801 skeletons received full
osteological analysis; "by far the largest Early-medieval burial population
so far recovered from Wales". They are the same excavation counted twice.
Please confirm which figure the HER now carries, and that both stages are
identified as such.

9. The c.1880 ironwork sentence, and what Amgueddfa Cymru now holds

GGAT 1990: "it is likely that some medieval ironwork which is held by the
National Museum of Wales was discovered under the rear wing c 1880 (RCAHMW
1988)." By 2005 that hedged sentence had become flat assertion in print.

Amgueddfa Cymru's disclosures of 25 August 2026 (FOI 2026-024 and internal
review 22 September 2026) and 22 September 2026 (FOI 2026-032) establish
the following, so far as the Museum holds recorded information:

(a) No accession, donation, transfer or deposit answering a visor, lance,
helmet, armour, soldier, horse, or nineteenth-century discovery under the
farmhouse floor.
(b) Accession 82.44H is the 1979 GGAT excavation assemblage. Collected by
Glamorgan-Gwent Archaeological Trust Ltd, 1979; donated 14 June 1982 by
Wales & West Housing Association Ltd; bibliographic reference Owen-John
1988. The collections database locality field is "Llandough Roman Villa",
NGR ST 168 733 — not "Great House Farm".
(c) Accession 95.56H is the 1994 Cotswold Archaeological Trust cemetery
assemblage. Collected by CAT, 1994; donated 28 July 1994 by Ideal Homes
Wales Ltd. Locality field: "Great House Farm Llandough", NGR ST 1681 7331.
Period on the human-remains entries: Early Medieval (c.400–c.1070).
(d) Accession 63.24 is objects stated to have been found at Church Farm,
Llandough, in 1858. The Museum's explanatory note of 22 September 2026
records that this is the only paperwork on that file, and that staff
consider it possible those objects are the basis of the later local story.
That is the Museum's observation, not a finding I ask you to adopt as fact.

Qualify or withdraw the 1880 sentence if you hold no source beyond the 1990
wording. Carry 82.44H and 95.56H on the principal HER description, with
collector, donor, donation date, locality field as catalogued, and NGR. If
the HER still points a professional user first to a post-medieval dwelling,
that is a record defect against what both excavations produced and what the
national museum now catalogues.

Please also confirm whether GGAT retained copies of the 1980 donation
correspondence with the National Museum (letters of 14 and 24 April 1980
are on the public record) and of any transfer file for 82.44H.

10. What the public record must now say

If GGAT02038s, or its successor identifier, still leads as a post-medieval
farm or dwelling, it does not satisfy the function of a working historic
environment record, and it is difficult to reconcile with section 194(2)(j)
of the Historic Environment (Wales) Act 2023. I have not seen the source
file, so I put that as a request for correction if the classification is as
described by Cadw officers, not as a finding I have verified on the face of
the record.

A professional user needs, in the principal description and not only in a
later narrative paragraph: the 1974 measured survey; GGAT's 1978–80 work
and 1982 advice; that the 1990 evaluation followed demolition and found
human bone; that the 1994 excavation recovered the assemblage above,
including in the 1990 low-potential zone; museum accessions 82.44H and
95.56H with their catalogued localities; and a note that the c.1880 NMW
ironwork attribution is uncorroborated by the institution said to hold the
object.

Please also reconcile the NGRs in use for the farmhouse, the 1990 trenches,
the 1994 excavation, the churchyard and the development parcels.
Distinguish this site from Great House Farm, Llandough near Cowbridge, and
from Llandow. Two Llandoughs sit in the same historic county. Homonym error
is a live records risk.

What I ask now

A. A complaint reference, the name of the person who will determine it, the
procedure you will apply, the timescale, and any independent review route.
Preserve all surviving Great House Farm records pending that determination.

B. Confirm which legal entity holds the records and functions formerly of
GGAT Ltd, and who handles HER corrections.

C. Answer sections 1 to 10 from the documents you hold. Where a document
does not survive, say when it left the file, under what retention rule, and
by whose authority.

D. Correct the HER so that the 1979 and 1994 investigations and accessions
82.44H and 95.56H sit in the principal description, the 1880 attribution is
qualified, the 1990 low-potential zone is annotated against the 1994
result, the three identifiers are reconciled, and Archwilio does not
silently substitute another trust's site.

E. Test and, if confirmed, correct the Archwilio substitution.

F. If you accept any part of this complaint, say what remedy you offer:
record correction, release of surviving files or an index, a written
account of the 1989–90 dual-role arrangement, and a methodological note on
the 1990 evaluation against the 1994 excavation.

I do not ask you to adjudicate title. I do ask you to account for what GGAT
knew, what it recorded, what it advised, what it then sold and executed,
what was found four years later, what the national museum now catalogues
from GGAT's own 1979 work, and whether the public record now tells the
truth of that sequence.

This letter is sent so that Heneb and GGAT Ltd have a fair opportunity to
investigate, answer, and offer remedy before any further process. Please
answer each numbered section. A general assurance that the organisation
acted appropriately will not do.

Regards,

Sion Buckler (Ret'd Sergeant)
Royal Signals Cyber/Electronic Warfare Professional and Life Member
[email redacted]

Please visit
greathousefarmwiki.wordpress.com

Enc: GGAT — Forensic Critique, Version 4.0.0, 21 September 2026
(to be attached before sending)