24 replies to this topic

[Jerry Bryan]

One of my New Year's resolution is not to whine so much on the RM forums. I made it through January 1st, but it appears that I won't be making it through January 2nd without whining at least a little bit.

I was processing some data for a Leander Levi Wheeler who was born 5 July 1859 and died 4 Apr 1941. These dates are confirmed by his grave marker which is available at www.findagrave.com. Another researcher reported that Leander was born in Long Creek, Macon County, Illinois but the other researcher didn't report a place of death.

There are lots of trees at ancestry.com which have data about Leander, mostly unsourced and probably copied from each other. The closest thing to real data on ancestry.com is a database called Illinois, Deaths and Stillbirths Index, 1916-1947. Even reading the description of this database, it's really hard to figure out where the data really came from. But mostly likely it came from indexing Illinois death certificates.

In any case, the Illinois, Deaths and Stillbirths Index, 1916-1947 database reports that Leander was born in Long Creek Twp, Macon County, Illinois and that he died in Long Creek Twp, Macon County, Illinois. As you can imagine, the subtle distinction between "Long Creek Twp, Macon County, Illinois" as reported by llinois, Deaths and Stillbirths Index, 1916-1947 and "Long Creek, Macon County, Illinois" as reported by another researcher caught my attention.

I seldom research Illinois data, and my ignorance of Macon County, Long Creek, Long Creek Twp, and the workings of the township system in Illinois is vast. So I didn't know if Leander was really born in the town of Long Creek, in Long Creek Twp, or what. And I didn't even know if the town of Long Creek was in Long Creek Twp or in some other township.

The RM gazetteer was actually more helpful that I expected. It included entries for Long Creek Twp, Macon, Illinois, United States and also for Long Creek, Macon, Illinois, United States. I was surprised to find the entry for Long Creek Twp because the Place Name Standard seems to have such an aversion to including the names of political subdivisions in standard names. A little further research indicated that Long Creek is a village (not a town) in Long Creek Twp in Macon County, Illinois. So it seems safe to assume that Leander was born in Long Creek Twp whether he was born inside the village of Long Creek or not. If he were born inside the township, then I might change his birth place to Long Creek, Long Creek Twp, Macon County, Illinois or I might just leave it as it is at Long Creek, Macon County, Illinois. But if he was born outside of the village, then I would change his birth place to Long Creek Twp, Macon County, Illinois. It's really impossible to tell what the other researcher meant.

Just to be clear, I'm whining much more about the Standard Place Name system than I am about RM. Standard Place Names shouldn't leave you dangling with ambiguities such as these. My whine about RM is simply that it's so committed to a naming standard that is such a poor standard.

Jerry

[Jimmy L]

Not to answer your question entirely, but as a native Illinoisan, the county, town or township system is generally quite simple.
The counties are all divided into townships, eve Cook County where Chicago lies. Townships have within them Villages and Cities which carry their own jurisdiction. Whether a place is referred to as XXXXXXX Twp, or Town of XXXXX, it is the same. Try not to confuse a village or other incorporated area with a town. Thus someone born in Elk Grove, Illinois before the Village of Elk Grove was incorporated, they were born in the Town of Elk Grove or Elk Grove Township. Most rural babies were not born in the village or city, but in the town.

[Dale DePriest]

It should be pointed out that this is not unique to Ill. One of the first tasks of the USA when it was founded as a country was to survey the land. This was done so that the government could sell it to its citizens to fund the running of the government. The survey system was divided into 6 mile by 6 mile squares called townships or simply towns. All the land was contained in one township or another. Even the census of today identifies some states populations by townships. See http://en.wikipedia....(United_States) for more details.

[Alfred]

Townships are a small divisions of a county. Here in South Dakota, most townships are six miles square, 36 sections of 640 acres each.
The Township board has the responsibility of taking care of the miner roads and a few other things.
There are Fedreal highways, state highways, county roads and township roads, which are mostly gravel surfaced.

Very few maps will show any distinction between townships, but if you live in the country you pay a bit of taxes to keep your township going.

Many townships have no town within them, so to indicate which part of a county the township is used.

[Jerry Bryan]

Much thanks to the enlightenment from several of you about towns and townships. I very vaguely knew how townships work (and I knew about the 6 mile x 6 mile squares), but I had no idea that in the context of the states with the township systems that the words "town" and "township" were more or less synonymous.

The 6 by 6 mile township surveys never really existed in the states I'm most familiar with such as Tennessee, Virginia, and North Carolina. And indeed, I'm most familiar with the term "town" being synonymous with "small city" and most of the "towns" I'm most familiar with actually are incorporated - apparently unlike in the township system. In the township system, apparently the villages are the incorporated entities, such as the Village of Long Creek in my example.

Given what you all have described, a "township" probably corresponds most similarly with a district in the states I'm most familiar with. I'll describe Tennessee, but I think North Carolina, Georgia, Alabama, and Texas - to name a few - work roughly the same way. When Tennessee became a state in 1796, taxes were paid to your county of residence but were paid through your local militia company. My sense is that military service of some sort through a local militia company was pretty universal, but I really don't know exactly how it worked. I've tried unsuccessfully for years to figure out the militia company system, and I've tried unsuccessfully for years to figure out which part of the various counties people lived in based on the militia company of which they were a member. Then in 1836 Tennessee created the district system which took effect in 1837. All taxes and deeds thereafter listed a district, which were probably about the same size as a township. However, the districts were not laid out in square grids. Rather, the district lines tend to follow creeks, rivers, ridge tops, etc. We are not flat here like in Illinois. Subsequently, I have found birth records, for example, that say that a person was born in District 9 of Anderson County, Tennessee. And all Tennessee death certificates since the beginning of death certificates list the district number (and similarly with Texas, Missouri, and several other states I research a lot).

In Tennessee, the districts are called Civil Districts. In other nearby states they are called such things as Judicial Districts, Magisterial Districts, and even Military Districts (perhaps harkening back to the old militia company system).

I was thinking a little bit more about my frustrating example of Long Creek (which in modern times, at least, is an incorporated village) vs. Long Creek Twp (which was and still is one of those 6 mile x 6 miles square things). It seems to me that a Standard Place Name has no trouble with either formulation, but that a Standard Place Name really can't handle both. Which is to say, it seems to me that a Standard Place Name can't handle Long Creek, Long Creek Twp, Macon, Illinois, United States because there is one too many items in the list.

If I understand Standard Place Names correctly, Long Creek, Macon, Illinois, United States and Long Creek Twp, Macon, Illinois, United States are both correct. There is nothing wrong with Long Creek Twp, Macon, Illinois, United States even though the information before the first comma includes three words and one of them is a place type identifier. To Standard Place Names, it appears that "Long Creek Twp" is just a name without any semantics implied by the standard, just like "Kansas City" is. A human being knows what "Twp" means (or at least, I do now) and a human being knows what "City" means, but a Standard Place Name doesn't have to know care. It's just part of the name of the place.

Finally, I can't stop before mentioning that in most states I'm familiar with that if you live in an incorporated area you still live in a county. But Virginia (and maybe Maryland, I'm not sure about that one) have a different system in which an incorporated area is not part of any county at all. I find that very strange.

Our friends outside of the U.S. must find all of this very mind boggling.

Jerry

[Laura]

RootsMagic doesn't care what a user enters for a place.

Any standards a user might follow are self imposed. Or maybe imposed upon them by how a online database or other genealogy program insists on how a place should be entered.

Even if you type in a place as the Gazzateer so RM will enter the co-ordinates for you, you can always then change the place to what you prefer.

I see no problem with a place name of City/Town/Village, Township, County, State, Country if that place name narrows the location of where the event took place.

In fact, if I had the land description of the place an event took place, I could enter the land desciption as part of the place. Meridian, Township, Section, Quarter, County, State, Country or maybe Lot, City, County, State, Country.

My purpose for entering places is to narrow down where an event happened to as close as my research gives me. I am not that concerned if that doesn't meet someone else's rigid standards. I am pretty good about imposing my own standards on myself.

[Candy]

I think Louisiana has a bit different, yet similar, system. Instead of having 'townships,' they have parishes, which are basically the same thing, I believe.

[Jerry Bryan]

RootsMagic doesn't care what a user enters for a place.

Actually it does care if you have County Check enabled. With County Check disabled, you are 100% correct.

I would love to enable County Check, but it's too painful to run with County Check enabled if I'm not following the Place Name Standard. I continue to believe that County Check could be rewritten slightly to be tolerant of place names that are not in 100% compliance with the Place Name Standard. The argument against enhancing County Check seems to be that doing so would make the users unhappy who like the way it works now.

Let's take my example. County Check is obviously very happy with Long Creek, Macon, Illinois, United States and with Long Creek Twp, Macon, Illinois, United States. The relevent data elements for County Check are "Macon", "Illinois", and the date of the event. There is no reason except for a little extra programming that County Check couldn't be equally as happy with Long Creek, Long Creek Twp, Macon County, Illinois. In the latter case, County Check again would have to determine that the relevant data elements are "Macon" and "Illinois", plus the date of the event. That doesn't seem like an especially onerous programming task.

Jerry

[Laura]

I can see where CountyCheck uses the usually accepted standard for entering county names since RM now syncs data to FamilySearch Family Tree.

I don't know how Family Tree would handle entering county names with County or Parrish in the place or places with more than city, county, state, country in the place.

At least, having CountyCheck brings the awareness to users of entering places as the place names for that date, not modern places.

CountyCheck only checks the county, state, and country with the date which user's usually put as the last 3 parts of a place. Whatever other parts of a place a user has, city, township, etc. before the county, state and country doesn't matter to CountyCheck.

I have CountyCheck disabled since I am not going to take County or Parrish out of my place names. If I ever decide to upload data from my database to Family Tree, I will do it from a copy of my database and make whatever changes are needed for Family Tree in the copied database.

I ran the Reports, Lists, CountyCheck, Errors report when CountyCheck was added to RM. I did have a few wrong places for the date I needed to correct, but it didn't give me errors for having County or Parrish in the Place name. I run the report occasionally when I have been adding dates and places as a check.

[Jerry Bryan]

I have spent the better part of the afternoon trying to chase down how all this Standard Place Name stuff really works to be sure I'm complaining about the correct issues. I have some questions and comments.

• I could have sworn that I found a document on the LDS Web site several years ago that defined their interpretation of the place name standard for genealogy. But today, I have not been able to find such a document on the LDS site or anywhere else.
• There seems to be a generally accepted place name standard that transcends genealogy. It seems to be endorsed by organizations such as the Census Bureau, the National Archives and Records Administration, and any number of other national and international standards organization.
• I can't seem to find a copy of this more encompassing place name standard, either. But based on blog posts and other messages that people post about it, it seems to be focused on the definition of the various data elements and their storage. It seems silent on such issues as commas or other punctuation in reports, the presence or absence of the word County in reports, the presence of absence of the name of the country in reports when the name of the country is clear from context. But not having been able to find the standard, I'm only guessing that it's really silent on issues of how a name should be displayed in reports.
• It appears to be that the standard is a four-level structure - country, First-Order Administrative Division, Second-Order Administrative Division, and Third-Order Administrative Division. In the United States, the First-Order Administrative Division would typically be a state, the Second-Order Administrative Division would typically be a county or parish, and the Third-Order Administrative Division would be anything else such as a city, town, village, district, etc. We typically display the Administrative Divisions smallest to largest rather than largest to smallest.
• I could certainly see cases where a genealogical report might need to have additional levels of Administrative Division. For example, I could see a possible need for something like Precinct 17, Knoxville, Knox, Tennessee, United States. In this case, United States is the country, Tennessee is the ADM1 (short for First-Order Administrative Division, in this case a state), Knox is the ADM2 (short for Second-Order Administrative Division, in this case a county), and Knoxville is the ADM3 (short for Third-Order Administrative Division, in this case a city). Then what is Precinct 17? This is where I really want to find the standard. But if I were working with a genealogy program where I could explicitly enter the country plus the ADM1, ADM2, and ADM3 data as their own separate fields and if there were no ADM4 field available, then I would simply enter ADM3 as "Precinct 17, Knoxville" without the quotes. Or depending on the situation with RM, I instead might enter the "Precinct 17" without the quotes as a Place Detail. Contrary to what all my whining about County Check through the years might suggest, I would have absolutely no problems whatsoever about storing my data in this fashion. Indeed, I would very much prefer that it be stored in this fashion than the way RM stores it now. I wouldn't mind storing Knox without the word County and I wouldn't mind storing the country explicitly as United States. Indeed (and repeating myself), I would prefer it.
• My problem with the Place Name Standard is the way it makes Place Names look in reports. This practically begs for Place Name templates that could control the way places look in reports. It would be a win-win situation and everybody would be happy - those who want to include the word County in reports and those who don't, and those who want to include the word "United States" reports and those who don't.
• But how would places look in GEDCOM? It should be possible to export places as raw ADM1, ADM2, ADM3, and country fields, or as an option to export them according to the same Place Name templates that are used for reports.
• Despite not being able to find an actual Place Name Standard document on the LDS Website, I did spend a while looking at StandardFinder at the LDS Web site (familysearch.org/stdfinder/). I had done so before, but it was a long time ago. I played around with putting places in to see what would come out. For one example, I gave it Oliver Springs, Tennessee. It gave me back three possibilities: Oliver Springs, Anderson, Tennessee, United States; Oliver Springs, Roane, Tennessee, United States; and Oliver Springs, Morgan, Tennessee, United States. In other words, it correctly recognized that Oliver Springs straddles three counties, but it wouldn't acknowledge the possibility that an event happened in Oliver Springs but that the county of the event might never be known. So I don't know how it wants me to enter just Oliver Springs, Tennessee.
• I was astounded to discover that StandardFinder converted Clare, Ireland into County Clare, Ireland. So StandardFinder likes the word County, provided only that you are in Ireland rather than in the U.S.
• Armed with this little bit of trivia about County Clare, I turned on County Check and played around a little bit with RM. Much to my surprise, County Check is accepting of both Clare, Ireland and County Clare, Ireland. It doesn't suggest to turn either one into the other.
• Armed with this little bit of surprise, I tried a few U.S. addresses. It seems that County Check is totally accepting of the presence or absence of the word County. What it's not accepting of is leaving out the country. So it will accept either Knox, Tennessee, United States or Knox County, Tennessee, United States. But it will accept neither of Knox, Tennessee nor Knox County, Tennessee. Is that the way it's always worked, or has something changed? If that's the way it's always worked, then for years I have been complaining about the wrong thing

Jerry

[Laura]

I have been playing with CountyCheck enabled today using US places.

CountyCheck has changed and is not poping up the CountyCheck screen when County is in the place name under some curcumstances.

I experimented without entering dates.

If I have just the county, state and country in the place name, I don't get the CountyCheck popup screen if I have County as part of the county name as long as I have entered the country spelled out as United States.

But if I enter the city or anything else in front of the County, I get the popup.

[Renee Zamora]

I asked the RootsMagician and he says we haven't make any changes to CountyCheck.

[John_of_Ross_County]

Coupled with the way of listing place names, there is a problem with changes of names. As an example, I have letter written in July 1841 from French Creek, Lewis County. This was in Virginia. About 1851, the county lines were changed and French Creek was then in Upshur County, Virginia, formed from part of Lewis County and a couple of other counties. This part of Virginia then became West Virginia on June 20, 1863.

I don't think there is any good way in RootsMagic to describe events in this community with a precise place name. Ideally, everything should be named as of the time the event took place. The problem comes when going to courthouses, cemeteries, or state archives in the present time. I would guess that 1841 land records from French Creek, Lewis County, Virginia, would be in the Lewis County Courthouse. But cemeteries once described as in French Creek, Lewis County, Virginia, would now be in Upshur County, West Virginia. I think the problem is that places named properly for the time frame when the event happened will sort alphabetically in different places when the name changes. Would some type of cross reference help? I am unlikely to visit this area in any case.

[Vyger]

Our friends outside of the U.S. must find all of this very mind boggling.

This has been discussed quite a few times over quite a few threads and recently here and I would hope that Rootsmagic has compiled the common user opinions and desires under one topic.

I did not know about the 6 mile by 6 mile Township thing but did read before that the Township system does not even extend over all the States of the US, also looking at the Township map of Wayne Co, PA things would not appear as clear cut as 6 miles by 6 miles with topography becoming a factor.

With so many land division differences throughout the world the only key here can be facilitating exceptions through County Check and also the Gazetteer which one would hope would overcome some of the objections to the current mode of operation. Ignoring the word "County" or "Co." within County Check is a simple to achieve solution but I would also like to see the facility for a user to build an add on Gazetteer file a bit like your custom spell check dictionary, I would doubt RM would want to facilitate the editing of the original

Users will always have to decide what should be PLACE and what should be PLACE DETAILS bit like Lumper or Splitter. This should be simple with Place Details being a building, street address etc but then all else should be recognized by the Gazetteer. Is the Cemetery the Place Detail or is the burial plot the Place Detail, I would say a combination of both.

As we go back with our genealogy and places it is obvious less will be recognized by the Gazetteer, County Check does some good work here (especially in nthe US) although with mixed results. Administrative districts were mentioned and historically in the UK and Ireland that administrative district was the Parish administered by the Church, even within these the Parish boundaries were subject to change Parishes were often split and the Roman Catholic and Protestant Parish boundaries were often different so we can see the scale of this problem.

Unique to Ireland was the Townland system which was a further sub division of land within the Parish, births were noted as taking place in a Townland, when a couple were married they were noted as being "of" a Townland.

With regard to County Clare, Ireland, County Check Explorer does not recognize variations of this place so I rather think County Checker is ignoring the entry as being unrecognized. One thing I would like to say using this example is that County Clare contains 81 parishes which contain 64 towns or villages and 2292 townlands, and a townland can range in size from 1 to several thousand acres. Very little below County level is recognized by the Gazetteer so would need to be manually geocoded until the "lift geocode from online map" becomes part of Rootsmagic. Regardless one could never expect Rootsmagic to contain such detailed land division information which returns again to my assertion that the program needs to facilitate custom entries which will change from Country to Country.

In Ireland an accurate description of an area of land in reverse order would be Country, Province, County, Parish, Townland although Province would be rarely used. Other land divisions used were "Poor Law Union" and "District Electoral Division". Electoral divisions are always changing and I pay little attention to these however the "Poor Law Union" was an administrative area for poor law relief from about 1840 well into the 20th century and I find it useful to have this as part of place notation for research purposes. Also, so as not to clutter the Place List beyond recognition I decided to record Townlands as Place Detail.

Now I strive to record historical Irish Places as Parish, Poor Law Union, County, Country with Townlands as Place Details within the Parish. When doing research in an unfamiliar area and having difficulty reading the place I can click on the Parish which I would know and then see a list of all the Townlands within that Parish, or search for the "Poor Law Union" (Rootsmagician, please add a search box to Place List). This helps me discern what the poor handwriting might refer to, helps me record it accurately and provides a sort of topographical reference list within Rootsmagic.

The above works well from a Mapping perspective as well which can be seen at the bottom of this page.

Modern place definition is completely different but as genealogy, by it's very essence, goes back in time I believe building these historical place histories and locations an important part of any research and also believe allowing user customizing here would be a winner for Rootsmagic when combined with the various other Place related wishes.

When I present my research to others I am invariably asked "...so exactly where was that?". Through the use of Notes, Media and Geocoding Rootsmagic provides me with the tools to have those locations properly recorded although it does not always make it easy to manage or report on.

[Jerry Bryan]

In Ireland an accurate description of an area of land in reverse order would be Country, Province, County, Parish, Townland although Province would be rarely used.

Just out of curiousity, in the standard "administrative level" system for geonaming (my term for it), would the proper correspondences in this case be country = Country, ADM1 = Province, ADM2 = County, and ADM3 = "Parish, Townland" (without the quotes). What I'm thinking is what if in Rootsmagic Version 17 some day, you could enter country, ADM1, ADM2, and ADM3 as four separate place fields that Rootsmagic would put together for you in reports as a single [Place] field. Would that work for you?

I think most typically with my limited knowledge of Irish geonames I would only know the country which I would enter as Ireland and the ADM2 which I would enter as Clare. I would love for [Place] in a sentence template therefore to be replaced by something like <[ADM1], ><County [ADM2], ><[ADM3], ><[Country]> from a proposed new place template if the place was in Ireland. Since I would only enter the Clare and the Ireland parts since that's all I would know, my [Place] would be come County Clare, Ireland. (As far as the place template itself, if this were for real I would have to think a little more deeply about the exact switch structure and where the commas would go with respect to the switches, but this gives the idea).

Jerry

[Vyger]

It does disappoint me that Rootsmagic appears to provide the smallest resource here and I believe years ago the Rootsmagician indicated that the Gazetteer would not be added to or improved. Legacy GeoDB (614MB on disc), FTM PlaceAuthority (333MB on disc) and Rootsmagic (140MB on disc).

I had another quick look at the rivals and I do very much like the option below which if it could be added to with custom user entries would provide an extremely useful resource. Another thing is that custom user databases for other Countries could be shared as resources by other users.

Whilst splitting is easy by delimiters to provide the hierarchical Place view much wished for splitting the Place Table into four separate fields may well have other very welcome implications and possibilities in the future.

I have tried the facility below on Russian Countys and various other locations and it works well so the information is out there and available.

[Vyger]

My apologies for taking Jerrys (again, again) thread on a slight aside but I do hope a new Place Management system includes the long overdue (IMO) Find option, Event Association window and a Live Map view which other software offerings now provide and have done for a while.

Personally I would drop RM Mapping and build an all encompassing Place Management UI around a 5 or 6 pane window.

[Laura]

I do not want RootsMagic to go back to having seperate boxes for parts of the place names as earlier versions of Family Origins and other genealogy programs had.

Genealogy programs changed to having one box for places with the parts of the place being separated with a comma when it became possible for good reasons.

One of them is the ease of typing in the place which was the user's number one complaint.
Enter City, Tab [or move cursor anorher way to the next box], enter County, Tab, enter State, Tab, enter Country and auto fill only works for the box you are in. And, every time you enter a place name for a fact, you have to go through the same process.

Entering a place into separate boxes takes much more time and effort than just typing the full place in one box. Doing column typing is much slower than doing straight typing where you don't have to move your fingers from the standard finger position.

And 4 boxes are not enough for many place names.

Then, we were also limited by how many characters could be entered into a box although that might not apply now.

I don't ecpect it to ever happen, but, if RootsMagic ever decides to go back to separate boxes for the parts of the place, I will be looking for a new genealogy program. And if all genealogy programs go back to that I will go back to paper.

[Renee Zamora]

Confirming discussion on this issue is noted in our tracking system.

[Vyger]

Genealogy programs changed to having one box for places with the parts of the place being separated with a comma when it became possible for good reasons.

One of them is the ease of typing in the place which was the user's number one complaint.

And when that happened users were completely free to type whatever freehand rubbish they wanted

It doesn't have to be one way or the other. a program could allow comma delimited input and provide a tool to reconcile that list in reverse order based of the comma delimited portions.