English speakers barely use the final consonant to tell these apart. They use the length of the vowel in front of it.
Final Devoicing, and Why It Hides This Pair
German and Korean both apply systematic final-consonant devoicing, where a voiced sound at the end of a word automatically turns voiceless. For speakers of either language these two words can be genuinely identical by ear, not merely hard to separate. The good news is that English does not really put the information there either. It puts it in the vowel, which your first language has no rule against hearing. You are not missing the cue, you have been listening in the wrong place.
The Answer Arrives Before the Consonant
Side by side the mechanism is obvious, and all of it happens before the final consonant is even reached.
bad
The vowel is noticeably stretched, and the final d is often barely released, sometimes you can hardly hear it. The length carries almost all the information.
bat
The vowel is clipped short and the final t is released crisply. Shorter vowel, sharper ending.
Listen Earlier Than Feels Right
This is the rule that makes ship and sheep unreliable when you judge them by duration, showing up from the other side: English stretches a vowel before a voiced consonant and shortens it before a voiceless one. Here that works in your favour, because it means the clue arrives before the sound you were straining to hear. If you are waiting for the end of the word to decide, you are listening too late, and the final consonant is the least reliable part of it anyway, since English speakers frequently swallow a final d. Move your attention onto the vowel and this pair gets easier immediately.
The Same Question at Other Ends
- ship / sheep
The same vowel-stretching rule, this time making a length-based guess unreliable.
- coat / goat
The same voiced-against-voiceless question at the start of a word instead of the end.
