Skip to main content
Log in

Getting (even more) serious about similarity

  • Published:
Biology & Philosophy Aims and scope Submit manuscript

Abstract

This paper critically examines Weisberg’s weighted feature matching account of model-world similarity. A number of concerns are raised, including that Weisberg provides an account of what underlies scientific judgments of relative similarity, when what is desired is an account of the sorts of model-target similarities that are necessary or sufficient for achieving particular types of modeling goal. Other concerns relate to the details of the account, in particular to the content of feature sets, the nature of shared features and the assumed independence of feature weightings.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Notes

  1. It is unclear why γ = 0 on Weisberg’s analysis here, as this seems to imply that it is irrelevant whether there are significant attributes of the target for which the mechanism(s) of the model cannot account.

  2. Other modeling goals discussed by Weisberg that call for different assignments to the weighting parameters include what he refers to as hyperaccurate modeling, where all of the parameters are assigned a value of one, minimal modeling, where ρ, α and β = 1 and θ, γ and δ = 0, and mechanistic modeling, where ρ, β and δ = 1 and θ, α and γ = 0 (see 2013, pp. 150–152 for further discussion).

  3. Weisberg acknowledges (2013, p. 150) an “ambiguity” in his discussion that relates closely to the conflation of aims identified here. However, it is unclear from his subsequent remarks whether he intends for his account to fulfill all three of the aims identified above or only one/some of them.

  4. Preliminary work along these lines also can be found in Contessa (2006), which argues that predictive modeling goals require a kind of structural similarity between models and their targets.

  5. This is in contrast to the features that Weisberg gives as examples—“oscillation”, “is a Lyaponov function”—none of which make reference to the target. See Weisberg (2013, p. 149).

References

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Wendy S. Parker.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Parker, W.S. Getting (even more) serious about similarity. Biol Philos 30, 267–276 (2015). https://doi.org/10.1007/s10539-013-9406-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10539-013-9406-y

Keywords

Navigation