Like many Americans, I descend from many families that have been traced fully in the literature, but I have also done personal research on a number of lines. It is possible to summarize the literature and to display my original research on the Internet, but for the latter a Web page is less than an ideal format. It is especially poorly adapted to the creation of a document that might in turn be cited as the source of a specific fact.
- URLs resist abbreviation. They cannot be shortened for brief reference and remain comprehensible to a computer.
When the data grow too extensive to fit on a single Web page, they must be connected to related data by either a link, a
chain of links, or a completely re-entered address every time. GEDCOMHTML conversions that place related facts in
different documents must be cited by a
completelydifferent URL with any reference other than a link from another HTML page in the same directory. Such references can never be concise enough for repeated use.
- When output is automated, file placement is usually volatile. Many of the same conversion programs that place related facts on different pages also shift them from one document to another whenever the conversion is repeated, producing the least reliable citations of any publication method. (The static media have their advantages.)
- Websites frequently have no internal division or organization. One may locate every instance of a specific word on a Web page with the Find function, but upon reopening the page there is no way to instantly recover the material of interest from within the page, short of Finding repeatedly. Even where targets are employednormally linked to subheadsthey rarely ease the processes of data recovery or brief citation. For all these reasons, references to HTML-rendered materials are often less exact than references to printed publications.
- HTML authoring tools cannot automatically correlate text with reference notes. Scholarly authors are generally forced to place references at the end of a web page (as I do whenever I think added commentary is in order) or to rely on procedures, such as framesets, that completely disassociate references from text. It is possible to target every reference note in HTML, but just as note numbers are likely to change with every modification of the text, the targets are too, just for ease of reading. HTML utilities generally range behind word processors in reference correlation, and references in HTML are entirely less convenient than in word processing software.
HTML is useful for very brief summaries of lineages found in printed works, as long as those publications are cited fully. I have also found it useful for summaries of research that may be controversial, can be explained by reference to a small number of core publications and documents, or is still at a preliminary stage. Generally, however, the challenges of maintaining scholarly standards when writing in HTML are imposing.
This website is organized around brief summaries of every distinct surname-lineage in HTML. On surnames that have required extensive documentary research, I also create PDF documents that fully explain the research and place references in convenient proximity to the text. I descend from some of these on multiple branches; others are unrelated to each other but bear the same surname. My methods for dealing with those situations are explained in more detail on the surname index pages, linked in the table at top left on this page.
From its beginnings in 2003, this project has been mostly limited to families that introduce themselves within twelve generations of my ancestry, counting myself as the first generation. I chose that limitation for practical reasons. On most of the ancestral lines that I have studied, twelve generations brings me to within one generation of immigrants to New England (the first population that I knew about in any depth), or their contemporaries in other regions. Research prior to that is beyond my expertise, andin cases where medieval descent is indicatedproduces pedigrees as complex as any to be found in America, implicating the entire medieval Western ruling class in the ancestry of a single seventeenth-century immigrant.
Even though I could expatiate on royal and gentry descent via at least two immigrants, I refrain from doing so here: Katherine Marbury, wife of Richard Scott, and Martha Bulkeley, wife of Abraham Mellows. Even with the limitation mentioned above, a twelve-generation span covers most of my colonial ancestry and a substantial proportion of my European ancestry. As I noted from the outset, my knowledge even of much of my nineteenth-century ancestry has never been complete, especially not in so far as it concerns nineteenth-century immigrants to the United States. I cannot and never could claim that these pages present my full ancestry for twelve generations or more. In a sense, this collection can only grow with additional research, even upon the progress that I and others have made over the past decade.
Austin W. Spencer | email: email@example.com