Working With Regional Obituaries For Local History Research
Most people who start digging through old Southwest Times obituaries run into the same wall within an afternoon. The microfilm is poorly indexed, the database scans are either too dark or cut off half the text, and the cross-references that actually matter are scattered across three different local archives. I spent about eighteen months piecing together migration patterns in the southern Southwest from nothing but death notices and funeral home ads. What follows is how I actually did it, not the version you find in a library guide. The first thing you need to understand is that standard obituary research is almost entirely useless for this particular newspaper if you're trying to reconstruct social history. A standard query for "John Smith died 1934" will get you a scan of one column. It won't tell you where he moved from, which church he attended, who survived him that isn't listed in the formal notice, or what the local economy looked like at the time of his death. You have to read past the death notice itself. I started by pulling the full range of issues from 1928 through 1952. That timeframe covers the tail end of the Dust Bowl displacement, the wartime factory boom, and the early suburban expansion. Most people stop at 1945. The stories in the later issues are where the demographic shifts become visible. I downloaded the digital archive but immediately realized the OCR was garbage on about forty percent of the pages. The typewriter-style fonts from the late thirties and early forties were barely readable by the scanning software. I stopped relying on text search after the first week and went back to manual page-by-page review for the core collection.
Here's what I did instead of searching. I pulled every obituary, yes, but I also pulled the accompanying content: the funeral home listings, the church announcement sections, the survivor notices filed separately from the main death announcement, and the real estate advertisements that appeared alongside the obituary blocks. A man dies in March 1941 and his wife's "in memoriam" notice runs two weeks later in a different column. The dates don't match. The location of the funeral home referenced in the second notice is different from the first. That discrepancy usually meant the family had relocated between the death and the final service, which is a migration event you'd miss entirely if you only looked at the primary obituary. The deeper work involved cross-referencing surnames across multiple years. I built a simple spreadsheet tracking every occurrence of a given surname in the obituary section over a five-year span. You start seeing patterns quickly. A family name appearing once in 1933 and then again in 1936 with a different first name and a new address usually indicates a death, a relocation, and a subsequent death of another family member at the new address. That's a migration chain. Sometimes it's a widow remarrying and changing her surname, which looks identical on the surface. You resolve the ambiguity by checking the wedding announcements in the society pages, which are in the same physical newspaper but filed under an entirely different section that most researchers ignore. I hit a specific problem with the 1943 to 1946 issues that nearly derailed the project. Paper rationing during the war meant the Southwest Times reduced obituary space significantly. Notices that would normally run two full columns were compressed into a single narrow column. More critically, the funeral home ads that normally provided the secondary details were dramatically reduced or replaced with generic religious content. I lost access to the survivor information that I usually rely on for family structure analysis. The workaround was to pull church bulletins from the same period. Several of the major denominations in the region published their own weekly notices that mirrored the funeral information the newspaper was no longer carrying. The public library had microfilm of those as well. It added about three weeks of work but recovered roughly sixty percent of the data I would have otherwise lost.
One counter-intuitive thing about these records: the most useful information is often in the errors. A transposed date, a misspelled street name, a incorrect age that turns out to match a census record when you adjust it. I found a cluster of three obituaries in 1931 where the age at death was consistently off by exactly ten years. The families were listing the wrong birth year because they were using the state's outdated baptismal records rather than actual birth certificates. Once I recognized the pattern, I adjusted the dataset and the migration timeline became coherent. Without that correction, the data suggested people were arriving in the region a decade earlier than they actually did. Another thing beginners miss is the distinction between the morning and evening editions. The Southwest Times ran both for several decades, and the obituary content differed between them. The morning edition carried the formal notices. The evening edition often ran expanded versions with additional surviving family members listed, community involvement details, and workplace affiliations. These appear under the same date but are physically separate publications. I learned this after wasting two days trying to reconcile inconsistent survivor lists. The morning and evening versions of the same obituary would list different numbers of surviving children. They weren't errors. They were different editorial choices for different editions. The digitization project for this newspaper is available through the state library system. The basic access is free with a state library card. There's a paid tier that offers higher-resolution scans and some OCR improvement, but the improvement is marginal on the pre-1940 material. I'd recommend the free tier unless you're working extensively with the late forties onward. The physical microfilm at the county historical society still has better image quality for the earliest issues, and the staff there can pull specific date ranges faster than you can navigate the digital interface during peak usage hours.
Get the Full Details
There are significant limitations to this approach that aren't worth softening. The obituary record is inherently biased toward people who could afford to publish. Wealthier families commissioned longer notices with more detail. Poorer families, particularly transient agricultural workers and Indigenous community members, are dramatically underrepresented or entirely absent. The obituaries you can find also skew toward white and English-speaking populations because the editorial decisions about who received notice were made by a small staff with clear biases. You'll reconstruct a partial picture of the region's history, and that picture will overrepresent certain communities while rendering others nearly invisible. Supplementing the obituaries with county health department mortality records and cemetery burial logs gets you closer to the full picture, but those records have their own gaps. Health department filings were inconsistent before 1940. Cemetery records depend on whether the burial ground was formally maintained or abandoned. Some of the older plots in rural areas have no recorded interments at all. The practical takeaway is that obituaries from this paper are a starting point, not a destination. They give you names, dates, and general locations. Everything beyond that requires corroboration from other sources. The cross-referencing work I described above is tedious and produces incomplete results even when done carefully. If you're patient and willing to spend time in physical archives rather than relying on the digital version, you'll find substantive historical material. If you need quick answers or complete data, this approach won't give it to you.
