Stock take of American research, May 2006

(new 21 May 2006)

Introduction

It’s a good time to do a stock take of what I’ve been doing with American records over the last few months. I've actually been "marking time" a bit on America recently while in discussion about the next phases of projects on the British side.

It would be great if this review opens up some contacts and opportunities for linkable American work progressing in parallel. I couldn’t transcribe everything single-handed and would be mad to try. A few common protocols are all that should be necessary for parts to be combinable in any of a variety of working arrangements to suit other personal agendas, needs and priorities, in connection with both existing and new projects. Very little computing knowledge is required beyond that involved in writing letters in any Windows-based text software (refer Spreadsheets for people who don’t already know that they already know how to use them). In linked items I’ve elaborated on the potential as I see it of the co-operative approaches particularly well-developed and widespread in genealogical circles (whose interests overlap with maritime history) to transform maritime history research resources (refer A Touch of Genie-us).

As I say in my Books, lists and computers discussion, it would be a pity if people who don’t relate to computers to the same degree see what I’m doing as irrelevant to them, or as relevant but something they can’t use or are too old to learn. If that sounds like you I hope you’ll read that discussion and try reading this item as simply relating to organising and distributing information. Even if you can’t do all the possible things with the results that you could if you were highly computer-literate you should be able to at least read and use the outputs in software you have on your pc and find ship names and numbers in them, just as you would in a book or a manuscript in Adobe Acrobat Reader. In my Spreadsheets feature I tell you how.

I’ve been working systematically on and off on a variety of American maritime history projects for more than a decade going back further to lists of ships on paper of the type that many readers will have compiled and still be working with. When I discovered what John Lyman and the Log Chips collaborators had been doing in the 1950’s I recognised immediately the same approach on a grander scale with the backing of a team (refer Log Chips) though Log Chips wound up as a journal while I was still in school and Lyman passed on almost 30 years ago.

Two strands of interest in particular drive my American projects; firstly, the development of sailing ship rigs and secondly, the application of basic descriptive statistics to provide national and global contexts for components of maritime history. That’s the order in which they developed but I’d put them the other way round now in terms of their importance and the contribution I can make. Most of the US information I’ve been collecting over some years relates to these but there are other starting points that can drive the process of defining maritime information needs and how computerised resources can best serve those needs. However, the nature of mine, particularly (2), means that what serves mine should serve many and in some degree at least perhaps all. My US square-riggers and the rig-size relationship feature demonstrates both of my driving interests, combining as it does, published statistics integrated with those compiled from individual ship records.

Over the last four years or so I’ve got deeply into issues of maritime information systems and databases (eg Ship identifiers in databases and indexes) arising out of reorganising and improving my initial attempts to computerise and extend my original paper records. This naturally and logically led into an appreciation of the value of ship official numbering systems (particularly the British and American as between them they cover half or more of the world’s merchant shipping over an extended period) and of Ian Buxton’s Global Ship Numbers project [http://gsn.ncl.ac.uk/ and GSN paper]. I’ll put a new item on my site elaborating these angles when I’ve worked through some of the issues further in private dialogue/s.

This will explain why the main thrust of my recent American work has been organising (and extending) my earlier records into and around a skeletal master index of American ships based on US official numbers. It cannot meet all research and database needs but can provide a major component outright that can be extended to accommodate the pre-1867 vessels (see concluding section of this item). My personal working files connect as subcomponents or linkable “add-ons” (as, in principle, could those of others and though such indexes to each other). Though my primary interest is sailing vessels, I am interested in steam and motor vessels through to the 1920’s as the context of sail and the transition from one to the other. Within that, I’ve paid particular attention to identifying steamers built outside the US as part of the interest in ship information systems in general and interlinking the national components.

A skeletal US master index – the concept

I have developed a rudimentary structure for such an index – so rudimentary that at this stage there is no expectation of attempting to include names and other information about every ship or even the majority of them – but which can be progressively developed (individually and collectively) to provide a basic index to ships and a master index pointing to where more substantial information about them can be obtained in various other sources.

If you visualise something that fulfils the functions both of a telephone directory and a tool for setting up telephone conference calls you’ll have the general idea. The critical point is that it provides a filing system of “pigeon holes” for each ship. It is infinitely extendable to index any sources whose contents can be tied to official numbers. It is technically possible to provide hyperlinks to more detailed data files designed to be used in conjunction with it. In principle, even hyperlinks to the internet itself can be included (though this last would be a mixed blessing unless one could be certain the url’s would remain unmodified indefinitely).

There are established British official number indexes equivalent to part (but only part) of this model that you can see on the Mariners' List at http://www.mariners-l.co.uk/ON1.htm and http://www.mariners-l.co.uk/IBON-INDEX.html. These have their origins in a limited goal of facilitating the use of British crew list records but an expanded form could serve wider objectives. American and British counterparts can be linked through those vessels which had histories under both jurisdictions.

You can visualise parallel American and British official number indexes as a capital letter "H" with the parallel uprights representing the two jurisdictions and the crossbar those vessels they have in common, all identified by one (or two) official numbers (and other identifiers). Several hundred ships that had both American and British official numbers are identified at http://www.mariners-l.co.uk/ON1.htm. I have identified some hundreds more but currently know generally only their American numbers. A ship search with just USA under Country [in which built] on the CHIN site at http://daryl.chin.gc.ca:8000/BASIS/vessel/public/vessel/sf will indicate upward of 2,300 Canadian-registered vessels alone that were built in the US, generally with their British (but not US) official number. (The CHIN search result will include some duplications and vessels that never had any official number but most will have had both. Any 7 digit numbers commencing with 9 in CHIN are surrogate numbers for vessels not allocated an official number or for which it is not known.) There will be more vessels registered under both jurisdictions at different times particularly after restrictions on the US registration of foreign-built ships that were nearly absolute in the 19^th century were relaxed. There was no such restriction on movement the other way in the relevant period.

A skeletal US master index – the specifics

Having developed my overall strategy with British records over more than a year, it took me only a couple of hours to compile the rudimentary first step of an American official number index. I deliberately used spreadsheet software that probably everybody reading this has on their pc even if they've never used it. The results are designed to be used with that software so that everybody can read them (see “Spreadsheets for people who don’t already know that they already know how to use them”) but with the advantage that it is elementary to import them into more sophisticated database software to serve complex types of need while still serving basic needs in universally accessible ways.

The index starting point is no more than lists of consecutive numbers representing each official number. Thanks to the unique way American official numbers were organised, it was half an hour's work to flesh out the bare numbers with columns to indicate the initial letter of the first name of the ship (up to 1903) and whether or not the number was reserved for an unrigged vessel (at least to WWII) - so immediately, even with nothing else at all, you have a tool providing some check on the reliability of information you have and perhaps a clue that a ship also had another identity you weren't aware of. (You'll find this information on a single page in the front of the List of Merchant Vessels of the United States from around 1903 onward; you don't have to get it second-hand from me.)

This needs to be progressively fleshed out as opportunities present with parallel columns to include any available other identifying numbers and codes of any possible information-linking utility, limited additional information about the ship with that number and indicating where additional information about that particular ship is available, hence the description as “a telephone directory and a tool for setting up conference line calls”. It is desirable that each number has more than just the number and preferably also a bit more than bare "name and address” for reasons of identification, error-prevention and to generate basic statistics, but provided the things being linked to it are accurate and follow some elementary rules of common coding and you know what you're doing then just the bare number and very little else will suffice for some purposes (as discussed in Access basics for combining and matching spreadsheets [forthcoming]).

Thousands of things that affect us all daily work in the same way from a unique identifying number to avoid ambiguity - my local newspaper's system for stopping and starting deliveries and managing the related company accounts works off the customers' landline telephone numbers because there is more than one way of writing many addresses and more than one person per household who could be making the call. Quite a lot of personal and institutional maritime resources do these things internally (which the user may or may not know - or need to know - is happening). In principle, it is elementary to link them externally systematically, amongst other ways with an index of the type I have outlined. Such an index can be extremely crude and simplistic in its most elementary form but the software permits it to be infinitely extendable in flexible variations to serve multiple research and organisational needs.

Indicating sources in an index

The US already has important maritime research resources that can be indexed within such a structure. A number do not have a direct counterpart on the British side. The List of Merchant Vessels of the United States as a major potential input has its British counterpart in The Mercantile Navy List.

I can already easily indicate in the skeleton index by their official numbers which vessels are already covered in the first (Sail 1867-1885) instalment of Jon Johansen's pioneering effort transcribing the List of Merchant Vessels of the US from 1867 onward. That information can then be used in two ways. You can work from the number in the index to whether or not Jon has transcribed the details for that vessel yet (and which of course automatically identifies it as a sailing vessel of the 1867-1885 period). You can work in the opposite direction from a reference in Jon’s CD to my index to see what it has to say about what else is also available for that ship elsewhere - initially the index may not indicate much else or anything at all but potentially much can be referenced in this way. It could well tell you the vessel was also owned in Canada and go to these places to see more information still when that aspect is developed.

In short order, I can extend the index to point the user to many other sources - pretty much anything else similar that Jon or anyone else produces, given only a list of the official numbers of the vessels in the source, in usable electronic form.

I have already developed the basics of a US signal codes' "subset" for inclusion in the index and/or as a linked add-on [refer American signal codes 1867-1932]. As a component of the whole, they are an approximate proxy for foreign-going vessels (generally exclusive of the Great Lakes). The US signal code lists contain some information not in the LMVUS standard tables and were generally published independently from 1895 on. Potentially, they can provide "quick and dirty" shortcuts to whether a vessel is still registered because the codes were allocated chronologically and are published consecutively when the (otherwise generally preferable) official numbers were not, so the two identifiers can be used in conjunction to advantage. American maritime historians cite signal codes to a much greater degree than their British counterparts, sometimes using them as an alternative to official numbers.

The Steamship Historical Society of America's "Lytle-Holdcamper List" of American steamers up to 1867 (refer Sources) is another very valuable resource that has no comparable international equivalent I know of. My skeletal master index could easily include another field/column giving the page number in Lytle-Holdcamper that will show where more information about the ship with that official number can be found. It makes no difference that it is a conventional book - the point is to highlight sources of further information whatever form it may take (sources do not have to be electronic though it helps if they are). Lytle-Holdcamper itself is a superior form of index. While Lytle-Holdcamper is a record of vessels registered up to 1867 (before there were US official numbers) many of these survived to be allocated official numbers which the book records. It would be only these vessels that would be page-indexed in the first instance. It isn’t a big job. I can do it readily enough but it would be an easily separable subproject that somebody else interested could do in collaboration freeing my time to develop some other aspect (write me at j_lowe@ihug.co.nz ).

[For the record - it is a completely separate matter not part of my present personal agenda - the print format of Lytle-Holdcamper should enable optical character recognition software to be used to turn it into a new electronic database much more easily and economically than most. That would of course involve an issue of copyright. If the Society or someone acting with their authority is interested I'll tell them what I know with some suggestions for design. More or less anyone could do the work who knows the minimum software basics.]

The Holdcamper index to New York pre-1867 registrations (refer Sources) is another valuable resource that must also include many vessels which survived 1867 to be allocated official numbers. However, it does not indicate which these vessels are or their official numbers (its primary function is to index the initial New York registration records which the user can use at National Archives and Records Administration to pick up the paper trail and follow it through successive re-registrations). Considerable work would be required to distinguish these and to add their official numbers but it would be worth the effort as it would provide a means to link post-1867 records to earlier port registration records. The addition would be done most efficiently in conjunction with an otherwise independent project to flesh out the limited information available for ships listed only in the first few years of the LMVUS (which does not include information on when and where built) and comparing tonnages under both the new and old measurement systems. There may also be copyright issues that would have to be addressed.

[With any necessary National Archives and Records Administration approval, it would also be technically straightforward to digitise the Holdcamper List using optical character recognition software. A pilot test indicates that it would be necessary to work from images obtained with a tripod-mounted digital camera, to disassemble and rebind a copy in order to scan complete pages on a flatbed scanner or to manually enter three or four complete lines on each page that would be lost in the “gutter” of the book when lain on a standard scanner.]

With the (feasible) addition of official numbers, a number of lists in “Log Chips” could be indexed in similar manner. I have already added most to the “Thober List” of square-riggers (although Thober needs a few corrections and quite a few additions).

There are more resources and there must be (I hope) some I haven't discovered yet (please write). The above will suffice to illustrate the principles and the type of opportunity and some obvious possible starters.

Apart from the straight indexing function I will be able to include limited identifying information from my own work in progress files for several thousand vessels to at least illustrate wider applications.

Basic information for more vessels would make the index altogether more readable and user-friendly and serve checking purposes. People who may wish to assist the development of the concept or see opportunities for promoting their own work by indicating its availability through my index may wish to consider releasing a few fields of information to enhance the index’s appeal and effectiveness in return for the potential benefits to them of the effect on the index’s circulation and greater utilisation enhancing awareness of the availability and utility of their complete databases and projects. It’s all matter of achieving “critical mass”. Combined effort can achieve what the individual cannot.

I will do quite a lot but I can't and won't attempt to transcribe hundreds of thousands of records personally - I would never do anything else and never finish so it would be a dead end (and the US isn’t my sole or principal field of interest). Nor need it be the life’s work of anybody else. That isn't necessary with major resources that already exist, which other people are working on or could work on independently or in conjunction. The "skeleton" index provides an infinitely flexible, extendable means for hooking up many information modules including ones yet to imagined by people I may never know. By providing a tool for linking and comparing different data files with overlapping content it provides a means for comparing alternative sources of the same information (and thereby a checking function) and for combining information from two or more sources to enhance both (refer Co-operative development of research resources).

Distribution of index

I have a “cunning” (?) strategy to make the skeletal index available inexpensively on CD-ROM for demonstration and some more substantial purposes (refer distribution on CD-ROM ).

I think the optimum time for release of “Version 1” would be when and if it could also serve to promote awareness of Jon Johansen's forthcoming second CD of LMVUS data. It could easily also index the page references to Lytle-Holdcamper and anything else anybody wants to offer in the meantime, supplemented with bits of my own work in progress. No attempt would be made at comprehensiveness but it should be possible to include enough content to make it interesting. A sufficient objective would be to demonstrate the utility of the approach and the opportunities it offers in a form that would permit it to take on an independent life of its own if need be.

I can add a field indicating the availability of information in any other source for which you can give me a list of the vessels' US official numbers in computer readable format (email me for specifics). The emphasis will be on the period 1867 through to the 1930’s but if you have anything outside that period that maybe could be integrated then let’s talk at j_lowe@ihug.co.nz ).

In principle, later versions would include enough additional content to assist people to add US official numbers to their own records and files that do not already have them in order to enable them to be linked by that means (the addition of official numbers to shipping arrivals files would be particularly advantageous). In principle, data files can be included on the same CD as the index for linkage to it by official number. In the first instance, the index would be limited to vessels that had a US official number and were therefore US-registered in 1867 or later. That would include many thousands built earlier (perhaps half those built in the US during the two preceding decades).

My web page (or some other by arrangement) would provide a link to fresh information and updates for any index distributed along such lines. Obviously if such an index were to establish itself as a useful tool for ongoing use it should find some native homes in the long run (plural would be my preference for survivability and no particular problem as it’s intended to be able to co-exist in multiple overlapping forms) though with modern technology it need not necessarily ever have a native base to fully serve American as well as international needs.

I'm under no illusions about changing the world or that the appeal of such things will be on a par with some flavour-of-the-moment pop star’s next song and dance routine, even within the maritime history community. I don't expect to start a movement, simply to provide a working prototype that can be developed if there is the interest and support. I won't be greatly disappointed or surprised if I don't get any feedback at all. The approach and the means of implementing it either connect or they don't. I'll launch my demonstration index in due course when time and tide permit and it can drift where it may, in Gulf Stream or doldrums, to be boarded or not at will. Email me at j_lowe@ihug.co.nz if you can see how it could serve you or you may have something to contribute.

US vessels before 1867

In all of these things the individual ship identifying number/s/codes are paramount in order to let the computer software do its job in the quickest and most efficient and reliable ways. (That’s actually true of even the paper-based counterparts as well.) Official numbers, where available, are ideal for the purpose. For those US ships that did not survive to be allocated an official number in 1867 it is more complex. There are means and a potential project that could provide a basis for giving them surrogate numbers as a by-product but that is academic for the moment. The present and foreseeable issue is the development of a database in relation to which the allocation of identifiers would actually have some significant meaning.

For vessels over 150 tons, the Rogers signal code list [Rogers] together with the intervening Commercial Code of Signals material [CCS] and other available resources have the potential to provide major components (but only components) of bridging from the start of American official numbers in 1867 back to the British beginning (1855) to "equal up the heights" of the two "uprights" of the “H” symbolising British and US official numbers in parallel. The US signal codes before 1867 differ from each other and from the 1867-1932 series – the earlier series are more useful for their ship listings than their actual codes.

A pre-1867 subset of vessels above 150 tons and foreign-going vessels could provide for many interests and needs for maybe less than five percent of the work needed to do all pre-1867 vessels - still quite a bit of work but it breaks down into bite-sized sub-projects (contact me if you are interested in working on any part).

Comprehensive coverage of all pre-1867 vessels would require building up the whole US, port by port, year by year - a much bigger project, though within the scale of genealogists’ team efforts. Holdcamper could provide part of New York (first NY registration only). The Works Progress Administration port registration transcriptions carried out in the 1930’s and 1940’s (refer WPA) obviously should be investigated as a potential primary source for some ports. There may well have been transcriptions done already that I don't know about (please let me know). Depending on the way they were typed and printed and the quality of the copy it may be possible to use optical character recognition software on them. That would have major implications for the practical aspects of transcription. Substantial manual checking would be required but not the double entry check desirable for manual transcription from scratch. I haven’t a clue about the copyright aspect.

feedback to j_lowe@ihug.co.nz

To return to main maritime index click Return To go back use Back arrow at top left of screen