Unnatural act for TMG

[Chris Heal 0]

Hello: My first post: I have been using TMG for perhaps 10 years, but not as an expert.

 

I'm doing some research into apprenticeships in the feltmaker trade in Bristol, England. These records start in 1535 and continue for about 400 years. Simply, a master takes an apprentice, perhaps, several apprentices. Some of these apprentices eventually go into trade and have apprentices themselves. Gradually, a tree of apprenticeship relationships develops.

 

The theory under test is whether apprenticeship trees lead to other business relationships which stay within a limited number of ' apprenticeship families'. These trees would be, in fact, the business networks, the trade's lifeblood. When partnerships are made, for instance, are they more likely to be constructed between tradesmen of the same tree even if they are only distant 'cousins'?

 

I decided to test whether I could use TMG to build apprenticeship trees with the master entered as the 'father' and his apprentices as 'sons', continuing the process as the 'sons' became 'fathers' to new apprentices. This is the 'unnatural act'. No other normal genealoical information is required except the year of apprenticeship. It is the basic relationship that is important, not the circumtances.

 

I soon got into difficulties. Individual apprentices were often 'turned over'. This means that the master died or left work or whatever and a new master was found. Two sequential masters for one apprentice is not unusual, nor three; the maximum I have found is six. How to give a son multiple masters - fathers? How best to identify the first master, second master, etc. I did this latter by adjusting the tag to, say, father master 1, 2 ... Then I couldn't get a decent descendant chart to display (this being one of the preferred outputs of the exercise). I got over this by 'marrying' each new masters for an apprentice to the first master. This worked for a small input, but when I got to the second generation, the chart was horrible with connecting lines going everywhere. A further complication is that, for instance, the second master may have himself been apprentice to the first master. And, just so it isn't easy, the first master was often the real father of the apprentice. Could this be indicated?

 

Is there anybody out there who finds this interesting and might have some ideas? At the moment, I am thinking that my 'unnatural act' is stretching TMG into places it just doesn't want to go.

 

Thank you

Chris Heal

[GenerationGoneBy 0]

If you want to show a chart type relationship, then you are headed in the right direction, but you have already seen the difficulties, because a CHILD can only have one primary FATHER.

 

You could also do this with tags, but then you wouldn't be able to chart out relationships.

 

I have just used tags for this, but I haven't tried to get relationships. For many of the apprentices I have found, the master became like a foster parent, and I created relationship tags for that, but the relationship is not primary. Since relationship tags can not have sentences, what you can do is create a tag that is used for reports (not charts) that states "John Jones became the master of Sam Smith on 23 Dec 1783, ..."

 

To get a nice chart, no I have no idea how to do it other than to have two people for each person, and link them to one person as a master and as another as a real parent. Even then you come across the issue of a person having more than one master.

[Michael Hannah 0]

Hi Chris,

 

I applaud you for "thinking outside the box" and setting up "pseudo" relationships, which is something I often do. And I am fascinated by your research as one branch of my family were apprentice chairmakers in Bristol. However, I think you have a level of complexity that may be beyond TMG. I am familiar with some "relationship" software packages that are designed to do what you are attempting, especially creating the charts you desire. Some of the best are used by intelligence and law enforcement but those organizations are using them to track criminal or foreign intelligence affiliations rather than apprentice feltmakers. The most widely used is a software product called "Analysts Notebook', but there are others. Even though I am one who often stretches TMG with "unnatural acts", with the quantity of people and the multiplicity of relationships you are trying to track I think you are stretching TMG a bit too far.

 

However, I have done a little experimenting with importing TMG data into such programs. The ability to set FLAGs in TMG with specific values to assign attributes for such people greatly enhances this process. Then using TMG's filters based on such FLAGs to output a List of People report in CSV format with their attribute flags is often an appropriate input format for such relationship programs.

 

Some ideas that may help in TMG. You might consider not trying to link the "real" people. Instead create one kind of "pseudo person" that was a "master" and a different kind of "pseudo person" that was an "apprentice" and use these "pseudo people" to construct your relationships and charts. A custom tag could link the "real" people to these "pseudo" people, but these pseudo people could be father/son. A "real" person that had multiple apprenticeships could give rise to multiple such "people" thus avoiding the one father/son relationship, but some kind of naming scheme might help you relate the multiple instances of apprentice "people" to the one "real" person. Hope this gives you ideas.

 

Good luck. Your efforts sound most interesting.

Edited April 27, 2009 by Michael Hannah

[Maccara 0]

A little late into this, but I've recently looked into some similar issues.

 

Analyst's Notebook & Sentinel Visualizer are ok for analyzing this kind of data, but as they are a bit cost prohibitive for me I looked into some alternatives and found out some interesting free (at least for non-commercial use) tools which can represent and/or analyze complex network data.

 

Something that can read GEDCOM directly is Pajek. Example of Pajek usage to produce classic p-graphs (and Petri nets) can be found here.

 

One very interesting tool for dynamic social network analysis is ORA and you might want to check out AutoMap as well (can be used to analyze narratives, for example).

 

Many of these tools can be used collaboratively (which is why I have chosen them) as you can convert/import/export data between them.

 

Much more tools and general information on social network analysis can be found at INSNA (check for example their software listings).