Reaching a well-weighed, informed conclusion is critical, even for an object as unassuming as an aged jug. If you want to understand something about the ancient Romans, you can’t simply talk to one, or view a photograph, or necessarily even read something he or she wrote—in essence, you completely lack direct and contemporary means of communication. You are heavily reliant the details of the community’s context and the bits you can glean from what remains of their milieu. Consequentially, one might observe and analyze the available artifacts (say, a grain jar), propose a hypothesis about how an object was used (perhaps for long-term storage in a state-centralized granary), and cast a discerning gaze on the other objects nearby to scrutinize whether they further support or undermine the idea. (Are the jars numerous? Is there evidence of accounting ledgers? Reliefs depicting the collecting and taxation of grains?) As an individual trained in the scientific method for my current means of employment, I find the process generally agreeable with the methodology of biological research: data collection, analysis, postulation of a hypothesis, and experimentation to yield more data that either support or refute your claim.
That isn’t to say, of course, that a conclusion is can’t be wrong. Sometimes, an object is modified over time, containing original materials that can be dated back to one era, and modifications in the style of a much later one. Forward five- or six-hundred years, and the form and functionality of the original object can be frustrating to uncouple from that of the modification, depending on how the adjustment was made. This kind of change, too, is an artifact, but in the other sense of the word: not only an object of human design, but also an artificial remnant of processing and handling.