Fig 1.
Basic architecture of the triangle model.
Fig 2.
Basic architecture of the Dual-Route Cascaded (DRC) model.
Fig 3.
Basic architecture of the Connectionist Dual Process (CDP) model.
Fig 4.
Basic architecture of the Naive Discriminative Reading Aloud (NDRA) model.
Fig 5.
The effect of length in word naming.
Fig 6.
Orthographic neighborhood density.
The effect of orthographic neighborhood density in word naming.
Fig 7.
Phonological neighborhood density.
The effect of phonological neighborhood density in word naming.
Fig 8.
The effect of body neighborhood density in word naming.
Table 1.
The linear interplay of orthographic, phonological and body neighborhood density.
Listed are t-values and β coefficients for each of the predictors in an additive linear model.
Fig 9.
Interplay of orthographic and phonological neighborhood density.
The non-linear interplay of orthographic neighborhood density and phonological neighborhood density in tensor product GAMs.
Fig 10.
Interplay of orthographic and body neighborhood density.
The non-linear interplay of orthographic neighborhood density and body neighborhood density in tensor product GAMs.
Fig 11.
Frequency by neighborhood density interaction.
The interaction of frequency with orthographic neighborhood density and phonological neighborhood density in tensor product GAMs.
Fig 12.
The effect of consistency of the orthography to phonology mapping in word naming.
Table 2.
The interplay of regularity and consistency.
Listed are t-values and β coefficients for each of the predictors in an additive linear model.
Fig 13.
Interplay of consistency and regularity.
The interplay of consistency and regularity in word naming. Top two rows shows results for regular words, bottom two rows for irregular words.
Fig 14.
Interplay of consistency and friends minus enemies.
The interaction of consistency with friends minus enemies in tensor product GAMs.
Fig 15.
Interplay of frequency and consistency.
The interaction of frequency with consistency in tensor product GAMs.
Fig 16.
The effect of frequency in word naming.
Fig 17.
The effect of familiarity in word naming.
Fig 18.
Mean bigram frequency and summed bigram frequency.
The effects of mean bigram frequency (top two rows) and summed bigram frequency (bottom two rows) in word naming.
Fig 19.
The effect of the frequency of the initial diphone in word naming.
Fig 20.
The effects of the number of simplex (top two rows) and complex (bottom two rows) synsets in word naming.
Fig 21.
The effect of family size in word naming.
Fig 22.
The effect of derivational entropy in word naming.
Fig 23.
Comparison of predictor coefficients for the observed data and the simulations of the ndra (top left panel), drc (top right panel), CDP+ (bottom left panel), and cdp++ (bottom right panel) models. Predictors from bottom to top: Freq (frequency), Orth (orthographic neighborhood density), FAM (familiarity), FS (family size), NCS (number of complex synsets), Phon (phonological neighborhood density), NSS (number of simplex synsets), DE (derivational entropy), REG (regularity), Cons (consistency), FID (frequency initial diphone), FE (friends-enemies measure), Body (body neighborhood density), BG (summed bigram frequency), BGM (mean bigram frequency), L (length).
Table 3.
Item-level correlations for the English Lexicon Project, Balota and Spieler (1998), Seidenberg and Waters (1989), Treiman (1995), and Kessler and Treiman (2002) data sets, as well as for a meta-analysis of these data sets.
Fig 24.
Quantile-quantile plots of the observed naming latencies and the naming latencies simulated by the ndra and cdp+ models.
Table 4.
Results of a principal components analysis on the 16 dimensional space described by the predictors.
Listed are predictor loadings for the first 8 principal components.
Table 5.
Results of a principal components analysis on the 16 dimensional space described by the predictors.
Listed are β coefficients for the first 8 principal components.
Table 6.
Results of a linear model predicting observed reaction times from model components.
Listed values are component t-values.
Fig 25.
The effect of frequency in non-word naming.
Fig 26.
Pronunciation performance: Non-words.
Non-word pronunciation performance for the nonwords from the ARC non-word database (left panel) and for the Pritchard et al. (2012) data (right panel). Lighter shaded green areas in both panels indicate additional pronunciation accuracy in the cdp+ model when the naming activation criterion parameter is changed from 0.67 to 0.50.