Focus article “All models are wrong but some are useful”: A response to Campbell's comment on estimating Mytilus californianus shell size Gerald G. Singh a , Iain McKechnie b, c, * , Todd J. Braje d , Breana Campbell d a Institute for Resources, Environment and Sustainability, University of British Columbia, Vancouver, BC, V6T 1Z4, Canada b Department of Anthropology,1218, University of Oregon, Eugene, OR, 97403, USA c Department of Archaeology, Hakai Institute, Simon Fraser University, Burnaby, BC, V5A 1S6, Canada d Department of Anthropology, San Diego State University, 5500 Campanile Dr., San Diego, CA, 92182-6040, USA article info Article history: Received 25 June 2015 Received in revised form 26 August 2015 Accepted 31 August 2015 Available online 3 September 2015 Keywords: Marine ecology Morphometric regression Historical ecology Zooarchaeology Size estimation Allometry Shellfish abstract Developing accurate methods for estimating animal body size from fragmentary remains is a key focus of zooarchaeological research. Here, we respond to Greg Campbell's critique regarding methods we recently developed to predict Mytilus californianus shell size from archaeological contexts using linear regression. We show that Campbell's assertion that our regressions are “inaccurate” is incorrect and mis- characterizes the premise and results of our study. We appreciate that Campbell draws attention to the importance of allometry but do not agree that archaeologists must first describe ontogenetic size re- lationships before developing a practical method for size prediction in zooarchaeology. We further argue that pooling data from broad geographic scales incorporates diverse growing conditions into a predictive model to account for the uncertainties across archaeological time scales. We conclude by highlighting the difference between zoological and zooarchaeological research goals and emphasize that the precision required for a particular analysis can create a mismatch between analytical expectations and archaeo- logically applicable research questions. © 2015 Elsevier Ltd. All rights reserved. 1. Introduction It is an accepted truism that “all models are wrong but some models are useful” (Box, 1978:2). In a series of papers, we devel- oped a method that provides a useful estimate of total mussel shell size from archaeological fragments (Campbell and Braje, 2015; McKechnie et al., 2015; Singh and McKechnie, 2015). Within the range of sizes considered, our models work well as shown by the fits of the data to the regression lines. Gregory Campbell's e henceforth GC's e comment (2015) focuses on what he considers to be faulty methods for estimating Mytilus californianus shell size. Unfortunately, GC's criticisms reflect mischaracterizations about the premise and results of our study. We find his narrowly focused comment worrisome, not just because he performs statistical procedures inappropriately but because of a mistaken belief that archaeologists must first describe the precise morphological growth relationships before developing a method for size prediction in zooarchaeology. Here, we respond to the “assump- tions” he presents in his comment and reflect more generally on differences between zoological and zooarchaeological approaches to size estimation. 1.1. GC's assumption #1: organisms' dimensions have straight-line relationships We do not dispute the importance of allometry (the relationship of body size and shape) as a foundational method for under- standing how organisms develop. While GC may be correct in his claim that the relationships between particular umbo measure- ments (and other measurements) and shell length are allometric, these specific relationships have not been found in the M. cal- ifornianus literature and our studies were not seeking to investigate whether this is the case. GC is right to note that Singh and McKechnie (2015) choose to present a non-linear allometric func- tion for meat-weight predictions. However, we did this because length-weight regressions have been clearly shown to follow this type of relationship across multiple mussel species (Rodhouse et al., 1984; Zotin and Ozernyuk, 2004). Yet with respect to shell size, isometric and even linear relationships have also been observed in * Corresponding author. E-mail addresses: geraldsingh@gmail.com (G.G. Singh), iainm@uoregon.edu (I. McKechnie), tbraje@mail.sdsu.edu (T.J. Braje), breana@scic.org (B. Campbell). Contents lists available at ScienceDirect Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas http://dx.doi.org/10.1016/j.jas.2015.08.021 0305-4403/© 2015 Elsevier Ltd. All rights reserved. Journal of Archaeological Science 63 (2015) 160e163