The authors used a case-series design to compare comprehension disorders of Wernicke’s aphasia (WA) and Semantic aphasia (SA) patients on Warrington’s paradigmatic assessment of semantic ‘access’ deficits. In this cyclical set of verbal (spoken word-picture) and non-verbal (picture-picture) matching tasks, a small set of semantically-related items are repeatedly presented over several cycles so that the target on one trial becomes a distractor on another (building up interference and eliciting semantic ‘blocking’ effects). The effect of lesion location was also investigated, separating WA and SA patients with damage restricted to temporoparietal regions from those with lesions encroaching into prefrontal areas. A multimodal semantic impairment was observed both in WA and in SA cases, although as expected the WA group showed greater deficits on auditory-verbal than picture judgements, with accuracy decreasing considerably as acoustic-phonological requirements increased. Different beneficial and harmful effects of stimulus repetition were observed. The former were related to aphasia classification, because WA cases showed initial improvement with repetition of words and pictures, while in SA, semantic access was initially good but declined in the face of competition from previous targets. The latter were predicted by lesion location, because the ability to re-select both word and picture targets in the face of mounting competition was linked to left prefrontal damage in both groups. These different effects were explained by assuming that the term ‘access disorder’ may refer to two somewhat different deficits: (a) an impaired entry from a particular modality (e.g., spoken words) into semantics, or (b) a deficit within the semantic system with regards to the activation/inhibition of stored semantic representations. The authors conclude that ‘refractory’ effects (i.e., negative effects of cycle) are comparable for words and pictures, not only in patients with SA, but also in patients with WA and that these effects are linked to damage to left prefrontal cortex, which has been previously associated with the control of competition from previously-relevant responses. WA patients have, therefore, two types of semantic access impairment : (a) difficulty with initial conceptual activation (ameliorated by repetition) and (b) difficulty in the face of strong competition (increased by repetition), while SA patients show the second type of semantic access deficit in isolation. Comments: the paper is beautiful, well constructed and constitutes a piece or a larger and more complex line of research, concerning the mechanisms and the neuroanatomical substrates of verbal and non-verbal semantic disturbances in aphasia. The only problem that I have concerns the need of giving a more detailed description of the criteria used to select WA patients to be included in the study, in order to avoid the danger of (wrongly) suggesting that most WA patients have the semantic disorders described in this paper. If we look at the inclusion criteria of Robson et al (2012a), cited in the methodological section, we see that they were required to show a comprehension impairment below the 45th percentile, but that in reality ‘all participants fell into the bottom 20th percentile on repetition and comprehension subtests’ and were, therefore, severe WA patients. I think that a similar detailed description of the characteristics of WA patients included in the present study could be useful to the reader, to delimitate the field of application of the reported results. Furthermore, the number of Tables could be slightly reduced.