Weighted parsimony outperforms other methods of phylogenetic inference under models appropriate for morphology.

Abstract in brief

• Weighted parsimony techniques outperform other phylogenetic inference methods when applied to morphological data, especially under realistic models of evolution.
• The controversy in phylogenetics centers around whether to rely on model-based approaches like likelihood and Bayesian methods, which require explicit statistical models, or on parsimony, which avoids assigning probabilities but offers a rational explanation of data.
• Recent simulations that favor model-based methods often assume DNA-like evolution for morphological traits, but this paper demonstrates that, under more realistic conditions, weighted parsimony is equally or more effective than these models.]