Nasals

Overview of Nasals

There are three nasal phonemes in English. They are the bilabial /m/, the alveolar /n/, and the velar /N/. We produce these nasal phonemes by lowering the velum to allow air to flow through the nasal cavity. Some languages such as French and Portuguese have nasalized vowels as phonemes; although such vowels do exists in English as allophones of vowel phonemes in the presence of nasal consonants, the three nasals listed above are consonants which involve complete obstruction of the oral cavity. The location of the obstruction determines the place of articulation of the nasal: the lips for /m/, the tip or blade of the tongue at the alveolar ridge for /n/, and the back of the tongue at the velum for /N/. Because of the complete obstruction of air in the oral cavity, nasals are sometimes called nasal stops.

Although air cannot flow out of the oral cavity during the production of a nasal, the sound waves within the oral cavity and those within the nasal cavity interact in complex ways. In addition to the resonances or poles which characterize the other sonorants, nasal sounds also have antiresonances or zeroes which affect the appearance of the formants in nasal phonemes. Remember that the liquid /l/ also has such antiresonances.

Spectral Cues

 

• Voiced but with lower amplitudes than the vowels or approximants.
• Formants spaced at average intervals of 850 Hz rather than 1000 Hz as in non-nasals because of the longer vocal tract.
• F1 in characteristic ``nasal murmur'' at 250 Hz.
• For non-intervocalic nasals, F1 often the only visible formant.
• Spectral discontinuities at the beginning and end of the nasal.
• Optional nasal release - looks like a weak plosive burst
• Nasalization in surrounding vowels (vowels with ``split'' F1).

Frequency levels for nasals

Keep in mind that the formant most affected by context is F2. Because of its sensitivity to place of articulation, F2 can be a strong cue to distinguish the three nasals both in the nasal segment itself and in surrounding phonemes. F1 and F3 are not as important. Remember that these values are target levels towards which F2 moves, not necessarily the frequency level which F2 actually attains.

• /m/ - F2 target level for bilabials - 900-1400 Hz
• /n/ - F2 target level for alveolars - 1650-1800 Hz
• /N/ - F2 target level for velars - 1900-2000 Hz

Spectrogram patterns for /m/, /n/, and /N/

• /m/
  The phoneme /m/ will have the typical bilabial cues: formants dipping from preceding sonorants into the /m/ and rising out of the /m/ into succeeding voiced phonemes. But with low back vowels such as /oU/ and /u/, F2 is already in its target position for /m/, so there will be no movement.
• /n/
  F2 moves toward 1650-1800 Hz if it is not already there. For /i:/ and /I/, this means downward movement; for /E/, F2 stays level. For /@/ and the back vowels, F2 moves up.
• /N/
  /N/ is easy to identify when it follows a high front vowel such as in the progressive suffix `ing' (/I N/) because of the ``velar pinch.'' Otherwise it is easy to mistake it for /n/. /N/ alters the appearance of some vowels quite substantially. /N/ cannot occur word initially. Speakers often insert a /g/ after the /N/, so look for a stop-like release.

  Figures 1 and 2 show some typical nasal phonemes in the spectrogram.