First, some of my linguistic background: I'm a native Cantonese Chinese speaker. I speak fluent Mandarin Chinese but with heavy Cantonese accent. I have a working-level proficiency in English, meaning I can communicate with native English speakers face-to-face but my speaking sounds very unnatural to native English ears.

As you know, in both Cantonese and Mandarin all plosives are unvoiced. We distinguish the plosives p, t, k vs b, d, g by aspiration. So the p, t, k are pronounced as [p^h], [t^h], [k^h] while the b, d, g are pronounced as [p], [t], [k] in Cantonese and Mandarin.

Now, I'm learning Portuguese as a very beginner. I was told that all plosives in Portuguese are unaspirated. The Portuguese distinguish the plosives p, t, k vs b, d, g by voicing. However, when I listen to a native Portuguese speaker pronouncing the p, t, k, I actually hear them as the aspirated unvoiced plosives [p^h], [t^h], [k^h].

Contrast with my experience learning Spanish. All plosives in Spanish are also unaspirated. I have a hard time distinguish the unvoiced plosives p, t, k vs the voiced plosives b, d, g in Spanish. They all sounds like b, d, g to me. Meanwhile, I can clearly distinguish the p, t, k vs the b, d, g in Portuguese. It's just that I hear the p, t, k as the aspirated plosives, even though I was told that Portuguese has no aspirated plosives!


  • 2
    Your observation tallies with my experience as well. As far as I know, it’s true that the phonemically distinguishing feature between /p t k/ and /b d g/ in Portuguese is voicing, not aspiration; but phonetically they are much more likely to be aspirated than in Spanish, though not normally as strongly as in English or German. Apr 26, 2023 at 11:20

1 Answer 1


There are two aspects to this:

  • the greater aspiration of /p, t, k/ in Portuguese than in Spanish.
  • the greater lenition of /b, d, g/ in Spanish than in Portuguese.

From one 2008 study of 35 South and SE Brazilians:

The first aspect to be observed is the presence of aspiration in BP voiceless stops. For velar and alveolar stops, the aspirated variant (allophone) is more frequent than the unaspirated one. For the bilabial stop, aspiration occurs in approximately 50 % of the samples.

The extent of the aspiration (more specifically, how positive the VOT is) is quite high for Brazilian Portuguese, measured at a mean of +47 ms for /kʰ/, but this is nowhere near the average for certain speakers of British English in a 2007 study, where /k/ is at +97 ms, nor for speakers of Mandarin Chinese across China, who average altogether +98 ms for /kʰ/ according to that study; and a separate 2010 study on HK Cantonese speakers had /kʰ/ above +100ms across all six tones.

Compare that to a 2020 study on five native Spanish speakers from Central Spain aged 52-71, where /k/ is at +21 ms, and the native British English speakers of similar age who had lived in Spain for more 10 years, whose /k/ in their pronunciation of Spanish was recorded at +28 ms.

The second aspect is probably more salient for native speakers of Portuguese and Spanish: Spanish lenites its voiced stops /b, d, g/ into fricatives /β, ð, ɣ/ pretty much everywhere except phrase initially and after a nasal consonant. /l/ also blocks /d/-lenition after it, e.g. aldehído, but permits /b/- and /g/-lenition in e.g. albóndiga and alguien.

Brazilian Portuguese generally does not have lenition of its voiced stops, but European Portuguese does do so (apart from phrase initially, lenition is also blocked after nasal vowels). This lenition is only really a feature of Northern and Central dialects within Portugal, though that does include the standard of Coimbra and Lisbon.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.