Talk:Mary Amelia Swift/GA1

GA Review
The edit link for this section can be used to add comments to the review.''

Reviewer: Mike Christie (talk · contribs) 01:36, 31 August 2022 (UTC)

I'll review this. Mike Christie (talk - contribs - library) 01:36, 31 August 2022 (UTC)

Images are appropriately tagged.
 * Two sources are unreliable or potentially unreliable -- see WP:RS/PS
 * ancestrylibrary.com -- this is an ancestry.com site
 * familysearch.com
 * This is a general comment, not aimed at you, Wikipedia editors typically have little understanding of the evidentiary process and in particular a misunderstanding of primary sourcing. (I spent a career analyzing evidence and being trained in evidentiary analysis and while that doesn't necessarily make me an expert, it does provide me with more understanding than the average Jane Doe. And yes, I totally understand that anyone can claim expertise without having it, so you'll just have to trust me and weigh my comments below.) Primary sourcing created by persons who have a vested interest in presenting material are very, very different than primary records created by say government entities or organizations which are merely creating records to document an event. With regard to historical persons and sources, verifying material presented from times when it was typical to "pretty up" history, requires evaluating primary sourcing which happened as near to an event as possible. If one relies solely on secondary sourcing, it can mean that one is merely repeating errors or interpretations, such as that statement by Allibone in this particular case, that were used to promote a person and their import. Neither of those websites are categorically unreliable and I would argue that they are no more unreliable than say google books. As with any evidence, when analyzing sourcing, one must 1) do due diligence to ensure that the source is referring to the topic in question and 2) evaluate whether it is likely to be reliable. In that regard, I have evaluated the sources, to wit
 * Banks: Both the author and the editor were employees of a government organization and had no apparent link to the subject, nor any discernible reason to fabricate or promote anyone in their presentation of the material. The publishers, i.e. a US government agency and a state library, are reliable organizations, independent of the subject and would not be likely to publish materials that were unreliable. Ancestry is merely acting as a host website for a published secondary source which is likely reliable. The information contained therein is a list of persons who are buried in a particular cemetery and who have been identified in other sources as family members. Meets both due diligence and reliability criteria.
 * 1850 census: The census taker, Conklin Smith was an employee of a government agency and had no apparent link to the subject, nor any discernible reason to fabricate or promote anyone in their presentation of the material. The publisher was a US government agency and would not likely publish materials that were unreliable, although there might have been errors because the persons involved fabricated information. It is a primary source, and not published until 70 years after it was taken per US law. The information contained therein is a list of persons who have been identified in other sources as family members. No OR, analysis, or conclusion is required to observe that the census taker listed the family as inhabitants of the 3rd ward of Brooklyn, or to see what he has written in the occupations. It is a limited use of a primary source and meets the criteria shown in WP:primary.
 * 1870 census: The census taker, Benjamin de Lamatis was an employee of a government agency and had no apparent link to the subject, nor any discernible reason to fabricate or promote anyone in their presentation of the material. Publisher rationale is the same as above for the 1850 census. The information contained therein is a list of persons who have been identified in other sources as family members. No OR, analysis, or conclusion is required to observe that the census taker listed her as an inhabitant of the 13th election district, 16th ward of New York City, or to see that he has written in the occupation for the children "at school". It is a limited use of a primary source and meets the WP criteria for including primary sourcing.
 * Connecticut Church Records: This is a published book of the records of a Congregational Church. The original authors were recording events, (birth, death, baptisms, membership, etc.) of their members and were unlikely to have fabricated the material presented or have had any direct ties to the subject. The editor and publisher of the material, is a state library, is a reliable organization, independent of the subject and would not be likely to publish materials that were unreliable or might damage their reputation. It is a published secondary source, though it was derived from primary sources. Ancestry.com is merely serving as the host website for a published work. The information contained therein is a statement that Dr. Swift's daughter was baptized on a particular date. No OR, analysis, or conclusion is required to restate those facts, thus it meets limited use of primary sources.
 * Will of Zephaniah Swift: It is a primary record, written by the judge to dispose of his assets upon death. As his motive would have been to define who was related to him and his heirs, we can be fairly certain that he was unlikely to fabricate that information. No OR, analysis, or conclusion is required to list the names of the persons who were designated in the will as beneficiaries, thus it meets WP's definition of limited use.
 * Marriage Record: It is a primary record created by a government clerk to record an event that established a new legal identity. It would be extremely unlikely that a government clerk would have fabricated such a document, thus it is deemed reliable. It does appear that I had originally concluded the family moved to Farmington at that time, rather than simply stating Harriet was born there as per the record. I have revised the statement to eliminate any analysis and state only that Harriet was born there. In so doing, it now meets WP's limited use criteria. SusunW (talk) 15:56, 31 August 2022 (UTC)


 * "based on her observances of teaching needs from her classroom experience": I think "observations" is the word needed here.
 * Thanks! Corrected. SusunW (talk) 17:04, 31 August 2022 (UTC)


 * I'm concerned about original research. It's OK for us to dig up newspaper articles that are not mentioned in primary sources and use them; it's not OK for us to synthesize information from them, which is what appears to be happening e.g. with Note 3, where we cite the 1838 newspaper clipping without knowing whether it refers to our Mary Swift.  I haven't traced the footnotes from Note 1 in enough detail to see if the same thing is happening there.
 * I have rewritten note 3 to eliminate comparisons of the sources. SusunW (talk) 17:04, 31 August 2022 (UTC)
 * With regard to Note 1, the analysis is wholly those of the authors, i.e. Reed says of Allibone "She is listed in Allibone's Dictionary as the daughter of Zephaniah Swift, L. L. D., 1759-1823. He was a Yale graduate, a lawyer, and at one time, Chief Justice of Connecticut. No dates are given for the daughter." (p. 154) Palmer analyzes in detail the biographical information in dictionaries and encyclopedias on page 170 and genealogical records on page 171. He states "The biographical references clearly record that Mary Swift was the daughter of Zephaniah Swift (1759–1823) whereas the genealogies equally clearly indicate that she was the daughter of Zephaniah Swift (1786–1834). This appears to be a situation that the methodology of triangulation does not solve. The widespread agreement of Swift family genealogists that Mary Swift was the daughter of Zephaniah Swift (1786–1834) makes the possibility that Allibone (1871, p. 2319) made a mistake and that later historians copied this error". (p. 172) Statements from Baldwin relay names of places the judge lived, which are supported by numerous other records. As stated above, the will was not analyzed, only the names of heirs listed. Statements that the doctor lived in western Connecticut and his widow and daughters survived him are restatements of information contained in sources. None of the material presented is original analysis. (Had that been the case, it could have been pointed out that the judge's will did not include a daughter named Mary and as a ten year old, she would not have been likely to have been disinherited and thus if he had such a daughter she was likely dead. Instead, the reviews of academics and basic facts were reported and it was left for the reader to draw their own conclusions.)
 * George Henry Swift stated that Mary was the daughter of the doctor and covered her biographical details. None of the information from him is OR on my part. Palmer evaluated the conflicting accounts and made the specific statement that Allibone may have been mistaken. The note is provided to present that there is another possibility for her parentage and the records conflict. In evaluating material, one makes a judgment on what is deemed to be a reliable source and explains discrepancies. In this case, the question is, is Swift or Allibone reliable? Based on the Oxford Reference description for Allibone's Critical Dictionary of English Literature was written with a goal to direct "the public to ‘the Best Works of the Best Authors’, and each book of any note is described with generous extracts from contemporary reviews and other critical writing", which IMO means it had promotional aims to make the subjects the "best" and claims should be scrutinized. Swift, on the other hand wrote to document his family record and in the introduction called for corrections and supplemental materials to be provided to him by readers, indicating to me that the goal was accuracy and it was likely reliable. Judging the reliability of a source is not the same as OR, it is merely analyzing whether a reference can be authoritative. (For the record, I wrote to professor Palmer and he responded. I asked him if the article accurately represented his analysis of Swift and if he had delved further into her biography. His reply was that "I read and enjoyed your Wikipedia article, which gives an excellent summary of Mary’s life and work. I am afraid that I have not done any further research on Mary Amelia Swift, but I hope to get back to looking at my old research again when I have more time".) I am open to discussion on how you think this might be presented better. SusunW (talk) 17:04, 31 August 2022 (UTC)


 * "and large typeface": "and a large typeface" would read more naturally, I think.
 * ✅ SusunW (talk) 17:04, 31 August 2022 (UTC)

-- Mike Christie (talk - contribs - library) 02:20, 31 August 2022 (UTC)
 * Thanks for picking this up I have responded to your queries and appreciate your thoughtful review. I am very interested in dialogue to confirm how the article can be improved. SusunW (talk) 17:04, 31 August 2022 (UTC)

Looking at the four familysearch citations, the issue is WP:PRIMARY; not that the source is reliable (sorry, should have taken more time and noted that in my initial comments). All four citations are used to support simple statements of fact, which is an acceptable use of primary sources, so I've struck that. Mike Christie (talk - contribs - library) 00:40, 1 September 2022 (UTC)
 * And a PS: not an issue for GA, but the source isn't really familysearch.org, it's the underlying document. You might consider restructuring the citation so that it points at the underlying source, with familysearch reduced to the role of a link. Mike Christie (talk - contribs -  library) 00:42, 1 September 2022 (UTC)

And now looking at the ancestrylibrary.com citations, I see they're hosted references that are in themselves reliable. I've struck my objection to those. Again, I owe you an apology -- I should have noticed the URL was just a link, and not sourcing to content generated at that site. Will look at your other comments next. Mike Christie (talk - contribs - library) 00:46, 1 September 2022 (UTC)

Passing. I think what you've done goes beyond what I would do myself, in that you're introducing records that we can't be certain are relevant -- I'm mostly thinking of the last part of note 3 here, and somewhat of the facts cited at the end of note 1. These are invitations to the reader to draw their own conclusions from the facts, which is more than I think we should be doing. However, I think this stays within the letter of the policies here, and my own preference for how to handle this sort of material isn't relevant, so I'm passing. Overall, in case it's not clear, I think this is an excellent article that presents an interesting story in straightforward prose. Mike Christie (talk - contribs - library) 01:26, 1 September 2022 (UTC)
 * FYI I added her to educators rather than scientists as that seemed more suitable. Mike Christie (talk - contribs - library) 01:29, 1 September 2022 (UTC)
 * , you don't owe me an apology. Our goal with GA whether writing it or reviewing it is to make articles the best that we can. Questioning sources and presentation is an important part of that analysis and had you not asked questions I would not have noticed the two instances where I probably had crossed a line into OR. I am unsure of how I could have formatted the sourcing differently, as for example on the census records the title is US census, publisher National Archives and Records Administration, etc. and the books also had all pertinent data entered. The mere input of the link to FamilySearch or Ancestry seems to trigger that red highlight and make it a visible warning. I am certain that I will never be able to effect change of the consensus that primary sources are inherently bad, so I just live with those horrible red cautions and when questioned, explain why I think whatever I used is indeed reliable. I truly appreciate your taking the time to weigh and consider why they were used. Educators is good. I agree that while hers was a science text, her goal was on education, not active research. I never quite know what category to nominate most women in. I thank you for helping to improve the article. SusunW (talk) 13:31, 1 September 2022 (UTC)