Speech is highly variable in rate, challenging the perception of sound contrasts that are dependent on duration. Listeners deal with such variability by perceiving incoming speech relative to the rate in the surrounding context. For instance, the same ambiguous vowel is more likely to be perceived as being long when embedded in a fast sentence, but as short when embedded in a slow sentence. However, it is still debated to what extent domain-general and domain-specific mechanisms (i.e., language- or speech-specific mechanisms) contribute to rate-dependent perception. Here we examined the role of domain-specific mechanisms in an implicit rate-normalization task in which we manipulated linguistic knowledge about how many syllables words have. Dutch participants were presented with lists of Dutch words that were acoustically ambiguous with regard to having one or two syllables (e.g., /k?ˈlɔm/ can be monosyllabic klom, /klɔm/, or bisyllabic kolom, /ko.ˈlɔm/). While being presented with these ambiguous word lists, they saw monosyllabic or bisyllabic transcriptions of the lists on the screen. We predicted that the same acoustic stimulus would be perceived as faster (more syllables per second) when combined with bisyllabic orthography compared to monosyllabic orthography. In turn, this would lead to downstream influences on vowel length perception in target words embedded within the word lists (rate-dependent perception of Dutch /ɑ/ vs./ /aː/). Despite evidence of successful orthographic disambiguation of the ambiguous word lists, we did not find evidence that linguistic knowledge influenced participants’ rate-dependent perception. Our results are best accounted for by a domain-general account of rate-dependent perception.