The Archaeological Record and Research Questions · Lesson 2

From broad interest to a testable question

7 min read

Lesson objectives

  • Explain From broad interest to a testable question with the evidence limits stated.
  • Produce a reviewable decision from fictional inputs.
Reading tools

“How did people live here?” is a worthwhile interest, but it does not yet tell a researcher what evidence to seek or what observation would change the account. A useful archaeological question identifies a phenomenon, a spatial and approximate temporal scope, and a comparison, then connects them to material traces that activities might leave. Consider: “Did use of the plain's edge change between phases A and B from food preparation toward pottery production?” The wording does not pretend that the phases have exact dates or that either activity excludes the other. It does identify testable expectations: plant remains and grinding tools on one account, production firing traces and malformed sherds on the other. A single indicator does not establish a site's function. Documented context and a competing explanation matter. Research frameworks organize current knowledge, gaps and questions, while a project design connects aims to appropriate methods. [Historic England, Research Frameworks, checked 2026-10-23; CIfA, Universal guidance for archaeological excavation, 2023, section 2]

Build the question through five entries: the units to compare, the measure, how it was recorded, a rival explanation, and the meaning of a non-detection. Comparing raw sherd counts from areas of different size confuses intensity with sample area. State a denominator, such as sherds per square metre, while also checking exposure, recovery and preservation. Likewise, an ash patch might reflect food preparation or a later fire. Look for evidence that discriminates between those accounts, such as the spatial relation of heat traces to malformed vessels, production waste, or food residues. Do not write one attractive hypothesis as though it were the result.

Make two hypotheses predict different patterns and specify what would weaken each. H1: pottery production in phase B left clustered firing traces and distorted ceramic pieces in related contexts. H2: household refuse or redeposited fill generated the burn and sherd distribution, so materials of mixed age occur without a fixed production focus. If sherds come from imported fill, or phase B has much better surface visibility, their abundance is not independent support for H1. State the limit of the design too: a few surface pieces cannot identify the makers or establish trade volume. The limit makes the eventual inference more useful because it tells a reader which decision the evidence can support.

You receive a budget that will cover only a small analytical sample. Before choosing samples, write down the decision the analysis could change. You may be asking whether production occurred, how two phases differ, or where transported material originated. Each requires a different comparison. If the question is production at this location, a more precise description of objects in imported fill may leave the central problem untouched. You first need a link between production waste and a stable surface. Sketch two decision paths: what would you recommend if waste connects to the surface, and what if it arrived with the fill? If both lead to the same recommendation, the proposed examination may have little relevance to the decision. This is a test of information usefulness rather than a precise financial valuation. It helps you select a consequential comparison and explain the choice to colleagues.

Lesson summary

“How did people live here?” is a worthwhile interest, but it does not yet tell a researcher what evidence to seek or what observation would change the account.