Toward a science of history.
Treat every record as countable data and test relations, whether the record is yesterday’s behavior or a hundred-year-old diary.
01Research in Context
What this study did
Funderburk et al. (1983) wrote a think-piece, not an experiment. They asked: can the story of the past become a real science like behavior analysis?
The team said historians should treat old letters, census sheets, and news clips as data points. Count what shows up again and again. Then test if one recorded event predicts the next, just like we test if attention predicts tantrums.
What they found
The paper does not give new numbers. It gives a recipe: stop guessing why things happened. Line up the records, turn them into counts, and run simple relations.
If historians follow this recipe, the past becomes a giant single-case data set we can analyze with graphs and stats.
How this fits with other research
Sidman (2024) shows the idea in action. He used fading and chaining data on the fly during early therapy sessions, proving a scientist-practitioner can decide in real time. Funderburk et al. (1983) wants historians to copy that same data-driven mindset.
Hatfield et al. (2019) hand us the modern toolkit—clear design rules, visual checks, and small-n stats. Those tools are exactly what the 1983 paper wished historians would adopt.
Virues-Ortega et al. (2021) describe Azrin’s loop: test, measure, tweak, publish fast. That loop mirrors the manipulative analysis Funderburk et al. (1983) says history needs. All three later papers extend the 1983 call into day-to-day behavior analysis.
Why it matters
You already graph behaviors and test functional relations. Funderburk et al. (1983) simply reminds us to treat any data pile the same way, even if it comes from 1920 court logs or last week’s session notes. When you write your next report, count the instances, plot them, and let the picture speak. That habit keeps our field a science, not a story.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick one past week of ABC data, count each antecedent type, and plot it against behavior frequency before you write the summary.
02At a glance
03Original abstract
The scientific status of History was compared to other sciences in the critical areas event selection, investigative operations, and theory construction. First, in terms of events studied, history is regarded as a quasi-scientific study of past events. However, viewed from the science of behavior's perspective of what historians actually do, history becomes a study of current records. As a study of currently existing records, not the non-existent past, history has potential to become a science. Second, like other scientists, historians may undertake manipulative investigations: they can locate the presence and absence of a condition in records and thereby determine its relation to other recorded phenomena. A limitation has been the lack of quantification that results from emphasis on the uniqueness of things rather than on their communality. Scientific training would facilitate viewing similar things as instances of a larger class that could be counted. Another limitation that cannot be easily overcome is the inability to produce raw data. This limitation has created problems in theoretical practices, the third area of comparison, because theoretical constructions have frequently been substituted for missing data. This problem too could be reduced through scientific training, particularly in other behavior sciences. An authentic science of history is possible.
The Behavior analyst, 1983 · doi:10.1007/BF03392391