London Lives - 1690 -1800. A new source of info

[buckler]

First the good news

Londonlives.org is now on line
By the makers of the fabulous Proceeding of the Old Bailey site (oldbaileyonline.com) this is a new site dedicated to providing an on- line searchable database of documents involving ordinary London people 1690 —1800 (actually 1680 —1820 in many cases). It includes the Old Bailey information, the PCC Wills data from the National Archives , plus a mass of new data sources. The new sources are a very mixed bag, probably based on what they could obtain, or had previously been digitised. Particularly useful I've found are the Polling Records and the Insurance Company Policy details although both have limited date ranges The insurance records are mostly new i.e. not on the A2A site records. A full list of the databases is given below .

SEARCHES include
Surname and Given Name by date range
Keyword by date range
They offer a Surname and Given Name search by occupation — but I cannot find how to work it. As not all names are tagged this is probably not worth much anyhow.


Now the bad news
The quality of some of the transcription is, frankly, very poor. They appear to have used at least some staff or subcontractors with no experience of either the period , London , the handwriting and simpe typing. This is a common problem on genealogical sites, but I'd expected rather better effective quality control from them. Where they provide a copy of the original its easy to check but a search for say "sausage maker " will not find sansage maker . And Drury Lane Tlidre should have been Drury Lane Theatre even to an unpractised eye, There are numerous typos also. The insurance records in particular are riddled with them. Also searching for “bucklemaker” will not find Fred Smith BKM as many of the voting records are coded.
The boolean search operators seen not to work - or maybe just not for me

The full list of their current files are , from their own list

Parish Archives
ï‚· St Botolph Aldgate
ï‚· St Clement Danes
ï‚· St Dionis Backchurch

Criminal Records
ï‚· Bridewell Royal Hospital
ï‚· Home Office
ï‚· Old Bailey Proceedings
ï‚· Old Bailey Sessions
ï‚· Ordinary's Accounts
ï‚· City of London Sessions
ï‚· Middlesex Sessions
ï‚· Westminster Sessions

Coroners' Records
ï‚· City of London Coroners
ï‚· Middlesex Coroners
ï‚· Westminster Coroners

Hospital and Guild Records
ï‚· Carpenters' Company
ï‚· St Thomas's Hospital

Additional Datasets
(most not a full run, I’ve added approx dates for some of the more interesting ones)

London/Westminster Directory 1700s
Salaries Paid to East India Clerks
4 Shillings in the Pound Tax, 1693-94
Income Tax Payments 1799-1802
Marine Society Boys, 1770-1873
Old Bailey Associated Records
PCC Abstracts of Wills
Fire Insurance Registers 1775 —1776 ,1780
Aldgate Marriage Duty Tax, 1695
Aldgate Parish Registers — 1685 - 1710 (some missing)
Aldgate Poll Tax, 1690-98
St Luke's Workhouse Registers
St Martin's Settlement Exams
St Martin's Workhouse Registers
Westminster Pollbooks 1749, 1774 ,1780, 1790,1802, 1818
Westminster Ratebooks.

These can be searched individually or by the full set.

VERDICT
Very good for what it's got, very promising.
But must try harder to reach the standard of the Old Bailey site.
Which although not perfect is as good as it is possible to expect
.

http://www.Londonlives.org
.

[dognose]

Hi Clive,

Thank you for posting this information.

Another source of detail regarding 18th century London will surely prove most useful and very welcome indeed.

Regards Trev.
.

[Tim Hitchcock]

Dear Clive, Thanks for drawing attention to the site. And thanks also for your critique, but I did want to explain some of the characteristics you criticise. First, the transcriptions are, as you note, marked by numerous transcription errors. They are probably only 97% accurate, but they are designed primarily as a finding aid, rather than as a scholarly edition, and are associated with an image of the original in all cases for the material we created ourselves. We assume that users will want to use the images as the basis for their own work. At the moment, becasue of the large number of 'additional datasets' covering the 1690s, these tend to swamp the first page of results, but the vast majority of the material is made up of transcribed text associated with an image of the original (240,000 images, and 40 million words). The transcriptions were created using the same methodology as the Old Bailey, by many of the same people (ie. they were 'double entry rekeyed' in India, before being 'tagged' in Britain). I suppose the only point I would like to make about this is that there is always a compromise to be reached between perfection and getting the material out there for others to use. And that whereas most sites hide their incredibly poor transcriptions (generally OCRd text, at an accuracy rate of 60-70% even for printed material), we are absolutely committed to showing the user what they are searching - exposing both the transcriptions and the XML tagging that contributes to the search facility (try out the 'view as XML' button at the bottom of the page). Most of this material is essentially ephemeral, and simply wont justify the cost of creating a fully accurate transcription. We are in the process of creating a user wiki that will allow people to log corrections - though I suspect that this will take a very long time to have a noticeable affect.

On searching for occupations - you need to register in order to access these search boxes. But, as you suggest, the level of tagging in the text does vary, and we believe that the keyword search facility (with variant spellings) is the most robust way in.

If the help text does not currently reflect the functionality in terms of Booliean operators and wild cards, please notify the site at: oldbailey@shef.ac.uk

On the additional datasets, these were, as you suggest, inherited from other projects, and reflect the standards applied by them. Ironically, the vast majority of them have been publicly available through the UK Data Archive and AHDS for decades (and online for the last ten years or so), but they were in awkward formats, and lacking in supporting documentation, and the wider scholarly community has not bothered to make any use of them. The intention here is to simply ensure that work that was originally undertaken in pursuit of free standing academic research projects is availble for re-use. We did work hard to expand the coding in cases where the original project provided an electronic guide, and to reformat databases where the inherited version was particularly difficult to use (the original of the Fire Insurance database included 140 coded fields). But many of the problems created by the original databases remain. If any of the interpretive pages included on the site are not accurate in terms of how they describe these resources, again, please let the project know.

Thanks again for the post.

All the best, Tim
.

[buckler]

Thanks very much to Tim Hitchcock, who is one of two project directors of the London Lives.

Firstly our thanks for the site itself, which has already produced serious information for me, and also for taking the trouble to explain some of the characteristics of the site in response to my posting. I hope my original comments did not upset him too much - as very clearly he is aware of some of the features [computer program speak for "minor bugs" ] .

Concerning the use of OCR on other sites I use the Times Digital Archive and their failure rate is far higher. A search on the Times site for a keyword one day will often produce the slightly different results on a subsequent occasion.

I'll do some testing of boolean operation and perhaps be able to give some "frinstances" .

One thing that the Polling and Insurance Records are showing is proof of what I have always suspected. The notification of change of addresses of silversmiths in in the Goldsmiths Hall Marks Registers are often delayed, in one case by over three years. I think that several unrecorded marks , and new marks used prior to apparent registration also suffer from the reluctance of silversmiths to visit Goldsmiths Hall on what they considered non-essential matters ! Human nature does not change.

Again our thanks to Tim and his team.
.

[buckler]

Tim's comments on Tagging not being complete and that a keyword search may produce more results is true even on the Person Name Search.

Walker entered into the SURNAME box and Brook into the GIVEN NAME box will return only two of Brook Walker's insurance policies. 1780 and 1786 . The 1780 giving him as a bucklemaker.

A search on the Keyword Search on "Brook Walker" will return three ! 1780, 1782 and 1786. As all three addresses are different this is useful information.

On the same record a search for bucklemaker and "buckle maker" will not find the 1780 entry - since Walker's occupation is actual entered on the record as buck lemaker .

Caveat researcha !

But nothing is perfect and this site as already produced worthwhile material for me.

As it expands so does our knowledge
.

[buckler]

I found a classic example of poor OCR not being picked up and cleaned up recently on this site .
Running a trace on Joseph Barker of Bread Street produced the utter gem

QUOTE

23 July 1714
GriffinLuke P dobeing charged by Joseph Barker
next the Olive tree upon Breed Street hill on a
Strong presumption of having pubed his porbett of a
handkercheife it being Seen to full from him & not
able to give any good account of himselfse
UNQUOTE
 
   " pubed his porbett" sounds like an unusual sexual offence rather than the action of a pickpocket !