Josephine Allen

PROOF CLASS: FATHER = ARGUMENT (Owen, 19 Sep 2026), MOTHER = UNDERDETERMINED (Owen, 19 Sep 2026) - 1870/1880/1900 say both her parents were England-born (K749, K750); that narrows it but names no woman, and no record calls Lydia her mother. PROOF SUMMARY (proof-standard.md 2b), written 19 Sep 2026, RESTATED the same day after three more records were read on the image. 1. THE QUESTION. Is [111] Josephine Allen (b. 21 Oct 1834 Baltimore, d. 16 Apr 1909 Decatur, Tex.) the daughter of William Allen and his wife Lydia of Washington borough, Washington Co., Pennsylvania? 2. THE EVIDENCE. (a) 1850 U.S. census, Washington boro, Washington Co., Pa., p.350b, dw 204 fam 223, READ ON THE IMAGE (K746): Josephine Allen, 16, b. MARYLAND, in the household of William Allen 63 b. ENGLAND and Lydia 52 b. England, with Mary 35 b. England, Ellen 21 b. England and William 18 b. Maryland. ESTABLISHES her presence, age, birth state and household in that borough in 1850. DOES NOT ESTABLISH any relationship - the 1850 schedule has no relationship column. (b) Texas Christian Advocate, 20 May 1909 p.14, obituary, read on the page image (C7039): "Mrs. Josephine Crouch (nee Allen) was born in Baltimore, Md., October 21, 1834; was married in Washington, Penn., to W. T. Crouch November 4, 1852". ESTABLISHES maiden name, birth date, birth city and the place of marriage. NAMES NO PARENTS. (c) Ancestry 1850 census index, three controlled searches, 19 Sep 2026 (K747): "Josephine Weldon b.1834 Maryland" returns 80 rows, ALL ENUMERATED - 0 named Josephine, 0 born Maryland - against a corpus control of 7,098 Maryland-born Weldons; and a surname-agnostic sweep of Washington Co., Pa. ranks this Josephine Allen FIRST with no Weldon in the county at all. (d) 1880 U.S. census, Arkadelphia, Clark Co., Ark., Enum. Dist. 33, p.466b, enumerated 8 Jun 1880, READ ON THE IMAGE (K749): on her OWN line - "Jose", 47, WIFE, b. Md. - the two parent columns read FATHER'S BIRTHPLACE ENGLAND, MOTHER'S BIRTHPLACE ENGLAND, written out in full while every other cell on the three household lines is an abbreviation. CONTROLLED: the same two columns were read for all three people in the household and give THREE DIFFERENT pairs (head Pa./Pa.; wife England/England; daughter Pa./Md.), and the daughter's pair is exactly what her parents state for themselves two lines above. That control was declared before the record was opened, because K484 and K495 both caught an 1880 informant writing one state repeatedly down a household. (e) 1900 U.S. census, Decatur, Wise Co., Tex., ward 1 sheet 2, dw 23 fam 24, READ ON THE IMAGE (K750): Josephine, Mother-in-law, b. Oct 1834, 65, married 47 years, 3 children 2 living, b. Maryland - FATHER'S BIRTHPLACE ENGLAND, MOTHER'S BIRTHPLACE ENGLAND. The same control, run again on FOUR people, gives FOUR DIFFERENT pairs (Germany/Germany, Pennsylvania/Maryland, Pennsylvania/Maryland, England/England), and her daughter Lizzie's pair again matches what her parents state for themselves. Her birth month and year here match (b) without being derived from it. (f) 1870 U.S. census, Arkadelphia, Clark Co., Ark., dw 37, p.202B, READ ON THE IMAGE (K750): the 1870 schedule asks not for a country but two yes/no columns, Father of foreign birth and Mother of foreign birth, which Ancestry does not index at all. BOTH ARE MARKED ON HER LINE and both are BLANK on her husband's and on both children's lines. The column is binary, so the mark's presence is its whole content; the letterforms were not resolved and nothing is claimed about them. (g) 1860 U.S. census, Arkadelphia, Clark Co., Ark., dw/fam 159, INDEX (K752): the household is W. T. Crouch 29, Josepine 26 b. Maryland, R. E. 7, W. B. 4, Mary E. 2/12 - and NO ALLEN. A widowed parent living with a married daughter is the commonest way a maiden surname enters a census; that route is a finished negative for this dwelling. 3. THE CORRELATION. (a) and (b) agree on four independent axes: the surname Allen, a birth year of c.1834, a birth state of Maryland, and residence in the very borough where she married two years later. (d), (e) and (f) add a fifth and different kind of agreement: her parents were FOREIGN-BORN, and specifically ENGLAND-born - which is what (a)'s household says of William and Lydia, and which the birthplace sequence of (a) already predicted (England, England, England, England, then MARYLAND for the two youngest, then Pennsylvania for the six-year-old: a family in Maryland at the right date). 4. THE CONFLICTS. (i) MAIDEN NAME. Two 1949 Wise Co., Tex. documents give her as "Josephine Weldon". They are ONE informant chain - a funeral record and the death certificate it fed - and (c) shows no such person exists in the 1850 index while this one does. Owen ruled 5 Aug 2026 to record both informants and change the name field for neither; THAT RULING STANDS and this note does not reopen it. (ii) LYDIA'S AGE, and this now cuts BOTH ways. At 52 (b. c.1798) she would have borne the Mary of 35 at about 17, which is the shape a first wife's children make - but she would have been about 31 at Ellen's birth, 34 at William's and 36 at Josephine's, all unremarkable. So the age structure argues AGAINST her being mother of the eldest and FOR her being mother of the youngest cluster, which is the cluster Josephine sits in. (iii) MARGARET RADCLIFF, 6, b. Pa., in the household under a different surname, is unexplained. 5. THE CONCLUSION. FATHER = ARGUMENT, and it is now INDEPENDENTLY CORROBORATED by the 1870, 1880 and 1900 censuses at (d), (e) and (f) above (K749, K750). This note recorded a falsification test on 19 Sep 2026 before the evidence was seen: the parent-birthplace columns of the 1880 Arkadelphia sheet were named as the thing that could overturn the grade. THAT TEST WAS RUN THE SAME DAY (K749) and returned England for the father - and the only England-born adult male in the household she was living in at 16 is William Allen, 63, b. England. The grade stands. MOTHER = UNDERDETERMINED, held deliberately after the new evidence, not for want of it. What the three censuses establish is that her mother was ENGLAND-BORN; what they do not do is name a woman. An England-born wife married to William AFTER October 1834 would satisfy every record read here and is excluded by nothing, and the 1850 household itself holds a second England-born woman, Mary, 35. Two further reasons to hold: the 1870, 1880 and 1900 statements are three records but may be ONE INFORMANT - Josephine herself - so the independent leg is still only (a); and NO SEARCH HAS EVER BEEN RUN for an earlier wife of William Allen, so the search is not reasonably exhaustive. Grading this now would grade the absence of a search. WHAT WOULD CLOSE THE MOTHER HALF: a record naming William Allen's wife before October 1834, or any record naming Josephine's mother. The best candidate WAS a Washington Co., Pa. probate of William Allen, findable from a candidate death of 26 Jun 1862; that candidate is REFUTED and the probate lane with it (K756, K757, K758) - the 1860 census shows his widow lodging in another household and he is not in the county, so he died 1850x1860, and a complete county sweep of the full-text probate series for 1858-1870 across five spellings of his name is empty against a denominator that proves those years are covered. The best candidate is now the First M.E. Church register at Washington, Pa. (Ancestry coll. 2451, HSP reel 427, c.1840-1866, 559 images, never browsed) - the congregation this family demonstrably belonged to - which keeps biennial death lists and may keep baptisms naming both parents. Then his naturalisation (he was England-born), or a Baltimore Methodist baptismal register c.1834-35, which would name both parents outright. WHAT WOULD OVERTURN THE FATHER HALF: a record placing a different Josephine Allen, b. Maryland c.1834, in Washington borough in 1850; or any record naming [111]'s father as someone else. Working papers: ancestry/sources/allen-111-1850-washington-borough-pa-2026-09-19/NOTES.md and ancestry/sources/allen-111-1880-arkadelphia-1900-decatur-2026-09-19/NOTES.md. ---- the note as it stood earlier on 19 Sep 2026 ---- ---- the note as it stood before 19 Sep 2026, with two corrections marked ---- AUDITED 5 Aug 2026 (supersedes the note of 4 Aug 2026) - [111] JOSEPHINE ALLEN. Her published obituary was found and read on the page image: Texas Christian Advocate (Dallas), 13 May 1909 p. 14, reprinted 20 May 1909 p. 14. In her own denomination's weekly it gives: maiden name ALLEN; born BALTIMORE, Md., 21 October 1834; married in Washington, Penn., 4 November 1852; removed to Arkansas 1855 and to Decatur, Texas 1895; died 16 April 1909 in the home of her son-in-law E. H. Baumgaertner. APPLIED FROM IT: birth tightened from the bare 'Oct 1834' to 21 Oct 1834, and birthplace from the bare 'Maryland' to Baltimore. The 16 Apr 1909 death already in the tree is confirmed exactly by the obituary. GRADE: STRONG, NOT PROVEN. An obituary is one informant, and here the informant is a grieving family reporting a birth 74 years earlier; 'born in Baltimore' is testimony, not a birth record. THE TWO PRINTINGS DIFFER ON THE HUSBAND'S INITIAL: 13 May 1909 reads 'W. F. Crouch', 20 May 1909 reads 'W. T. Crouch'. Both images read at zoom. The reprint is the correction and it runs toward the tree's William T. Crouch [110]. ✪ MARRIAGE DATE - RESOLVED 5 Aug 2026, superseding the 4 Aug paragraph that said the field had NOT been changed. Owen ruled: adopt the obituary's 4 November 1852 over the tree's uncited '1853'. The F71 marriage event now reads 4 Nov 1852 and carries citation C7039. The reasoning is not that 1852 is likelier but that a CITED date is checkable and an uncited one is inert. NOT A CONFLICT IF THE RETURN SURFACES: a ceremony on 4 Nov 1852 and a county RETURN recorded in 1853 are compatible - returns were routinely filed weeks or months later, and 4 November sits close enough to year-end for a filing to slip. If a Washington Co., Pa. return ever appears dated 1853 it RECONCILES both values rather than refuting the obituary. ⛔ THE NAMED NEXT DOCUMENT IS DEAD - a finished negative, 5 Aug 2026. There is no Washington County, Pennsylvania marriage register for 1852-54. The Pennsylvania State Archives RG-47 series list gives that county one 1852-era vital series, an Index to Birth Records 1852-1853; its marriage series begin 1885. Control: 26 of the 62 counties on the same page DO carry an 1852-55 marriage series, so this is a county-specific loss, not a state-wide absence and not an instrument failure. The one published transcript (E. B. Iams, copied pre-1944) notes "The originals do not seem to be in existance anymore" - and it DOES contain a different Crouch marriage six weeks later at Washington, Pa. (Watson Crouch, 14 Dec 1852), which is the control proving the corpus would have shown ours had it survived. Stop queuing this document. ⛔ HER OWN TEXAS DEATH CERTIFICATE CANNOT NAME HER PARENTS EITHER: Wise Co. file 990, retrieved and read - father and mother fields BLANK. ⛔ HER MAIDEN NAME IS IN DISPUTE, AND THE NAME FIELD HAS DELIBERATELY NOT BEEN CHANGED. Two Wise County documents give it as WELDON, not Allen: the 1949 Texas death certificate of her daughter Lizzie C. Crouch Baumgaertner ("MOTHER'S MAIDEN NAME - Josephine Weldon - Maryland") and Christian Funeral Home Record of Funeral no. 2593 for the same death. These are very probably ONE 1949 informant chain, not two independent informants - a funeral record and the certificate it fed - so this is one 1949 informant against the 1909 obituary's "nee Allen", not two against one. Both agree on Maryland. A fourth reading fits everything: that Allen was her MIDDLE name and Weldon her surname - her headstone reads "JOSEPHINE A." Owen ruled 5 Aug 2026: record both informants, change nothing. ☞ CONSEQUENCE FOR SEARCHING: every future census hunt for her must be run under BOTH surnames. The decisive documents - 1850 (FOUND 19 Sep 2026, and NOT in Baltimore: Washington boro, Washington Co., Pa., K746), 1860 Clark Co. Ark., 1900 Decatur - are all sign-in-walled and are browser tasks. PARENTAGE, 19 Sep 2026: her FATHER is now in the tree - William Allen of Washington borough, Washington Co., Pa. - graded ARGUMENT; see the PROOF CLASS head at the top of this note. Her mother is not flagged and is still owed. (This paragraph previously recorded that no parent was known.) INSTRUMENT NOTE: the county paper could never have answered this - the Portal to Texas History holds zero Decatur News issues for 1902 and zero for 1909. The denominational weekly did. For a Methodist Texan c. 1850s-1920s, search the denominational weekly before the county weekly.