The main focus of my talk will be the question “How can we ascribe notions from theories of philosophy of science to past scientists (in my research: 17th-century natural philosophers) who clearly lacked such notions?”. Such an enterprise seems to be vexed with the principle of accessibility raised by the anti-anachronistic critiques by Quentin Skinner and G.E.R. Lloyd (Skinner, 1969; Lloyd, 1990). Ergo, a historian needs to stick to the socio-historical standards, categories and concepts that were accessible to the historical figure one studies. Transcending this boundary is equivalent to anachronism. Ideas are only accessible from within the socio-historical framework they are formulated in. I my talk I wish to show that the situation is not alarming. An adequate position is one where difference-similarity between ideas is dependent on the stringency of our intellectual aims. If differences are irrelevant for our pragmatic purposes, they should not be treated. If our criteria are rather loose we may in a certain context ascertain a similarity, while in another more stringent context we may say that there are differences. This depends on the criteria that are relevant in view of the type of questions we pose. Similarity between ideas is a “triadic” relation: there is a similarity between two ideas in view of a set of criteria (hence: “x is similar to y with respect to z”). Difference-similarity between relations are a different matter. This relation is rather dyadic: a set of relation is similar to another set of relations when they share the same structure. Along these lines I attempt to make these matters more explicit. Parallels, i.e. partial identities, should not be confused with (complete) identities. When we say that there is a partial identity (PI) between an idea x and idea y means that there is a relevant set of characteristics (SRC) that pertains to x as well as y: Partial Identity (PI): PI(x, y) is true iff : SRC(x) = SRC(y) Not all the characteristics have to be identical. One has to decide which characteristics are relevant for a certain notion (SRC(x)). This notion will be dependent of the intellectual goals we have. In my talk I will present some examples. Then I introduce to types of predicates we can ascribe to past thinkers. In the first case, we attribute a relation of partial identity between certain modern ideas and past ideas (I-predicates). In the second, we attribute a relation of local identity between the structure of the relationships between past ideas and the structure of the relationships between modern ideas (R-predicates). I characterize the first predicates as follows: Partial Identity of Ideas (PII): PII SRC(Ipast, Ipresent) is true iff: (PII SRC(Ipast, Ipresent) reads as: there is partial identity between an idea from the past and a present idea with respect to a set of relevant characteristics) SRC(Ipast) = SRC(Ipresent) The second type of predicates is characterized as follows: Local Identity of Structure (LIS) LIS(S(Ipast), S(Ipresent)) is true iff: (LIS(S(Ipast), S(Ipresent) reads as: there is local identity of structure between a set of ideas from the past and a similar set of ideas in the present) S(RIpast) = S(RIpresent) I will discuss this distinction by means of some examples from Galileo’s Discorsi (1638). These examples hopefully make clear that we can indeed “derive enlightenment from looking at their work with clearer concepts” (Shea, 1983: 17). References: GALILEI, Galileo (1954), Dialogues Concerning Two New Sciences, translated by Henry Crew and Alfonso de Salvio, Dover: New York. LLOYD, G.E.R. (1990). Demystifying Mentalities. Cambridge University Press: Cambridge/New York/Melbourne/Sydney. SHEA, William (1983). “Do Historians and Philosophers of Science Share the Same Heritage?”, in: Shea, William (Ed.), Nature Mathematized, Historical and Philosophical Studies in Classical Modern Natural Philosophy (pp.3-20), Reidel: Dordrecht/Boston/London. SKINNER, Quentin (1969). “Meaning and Understanding in the History of Ideas”, History and Theory, 8 (1), pp.3-53.