CONNECTED SPEECH
Definition
Connected speech refers to the way sounds change, merge, and transform when people speak naturally and fluently, rather than pronouncing each word in isolation. In real conversation, the boundary between words blurs. Sounds at the end of one word blend into the beginning of the next, creating a smooth, continuous stream of sound. This is one of the primary reasons native speech is so difficult for learners to parse — the words you studied individually no longer sound the same when spoken together.
Connected speech is not ‘lazy’ or ‘careless’ speech. It is the natural, efficient way the human mouth produces language at normal conversational speed. Every native speaker of every language uses connected speech. In American English, it is particularly prominent because American speakers tend to speak at a moderate-to-fast pace with a relaxed, flowing style.
Key Rules of Connected Speech
Rule 1: LINKING (Consonant to Vowel): When a word ends in a consonant sound and the next word begins with a vowel sound, the final consonant attaches itself to the beginning of the next word. The phrase ‘turn it off’ sounds like ‘tur-nit-off’.
Rule 2: ASSIMILATION: A sound at the end of a word changes to match or blend with the first sound of the next word. ‘Did you’ becomes ‘didja’ as the /d/ and /y/ merge into /dʒ/.
Rule 3: ELISION (Sound Dropping): In fast speech, sounds that are difficult to articulate in sequence are simply dropped. ‘Next day’ becomes ‘nex’ day,’ ‘most people’ becomes ‘mos’ people.’
Rule 4: INTRUSION: A new sound is added between two vowel sounds to help the mouth transition smoothly. ‘Go on’ can sound like ‘gow-won,’ with a /w/ inserted.
Rule 5: GEMINATION (Double Consonants): When the same consonant ends one word and begins the next, the sound is held slightly longer rather than said twice. ‘Black cat’ sounds like one long /k/ sound, not two separate ones.
Examples in American English
Example 1: ‘I want to eat a apple’ → sounds like ‘I wanna eet-uh apple.’ The /t/ at the end of ‘eat’ links to the /uh/ beginning of ‘a,’ and ‘want to’ reduces to ‘wanna.’
Example 2: ‘What are you doing?’ → sounds like ‘Whaddaya doin’?’ Multiple assimilations and elisions combine to transform this common question dramatically.
Example 3: ‘He called us up’ → the /d/ at the end of ‘called’ links with /ʌ/ in ‘us,’ making it sound like ‘He call-dus up.’ The /d/ appears to move to the beginning of ‘us.’
Example 4: ‘Last night’ → the /t/ in ‘last’ is dropped, producing ‘las’ night.’ This elision is extremely common before consonants in American English.
Example 5: ‘She has a great idea’ → sounds like ‘She haza great-uh idea.’ The /z/ in ‘has’ links to ‘a,’ and the intrusive /uh/ appears between ‘great’ and ‘idea’ to bridge the vowel sounds.
Extended Dialogue: Connected Speech in Action
The following dialogue between two coworkers — Marcus and Brianna — leaving work on a Friday is full of connected speech. Read the ‘Written Form’ first, then study the ‘Sounds Like’ column to see how native speakers would actually produce these phrases. The annotations in brackets explain each phenomenon.
Scene: Office hallway, end of the workday, Friday afternoon
Marcus: Hey, did you finish up that report for Anderson?
[‘did you’ → ‘didja’; ‘finish up’ → ‘finish-up’ (linked)]
Brianna: I did, but it took me all night last night to get it done.
[‘last night’ → ‘las’ night’ (elision of /t/); ‘get it done’ → ‘geddit done’ (linking + flap T)]
Marcus: No kidding! I thought Anderson sent out the deadline extension?
[‘No kidding’ → ‘No kiddin” (elision); ‘sent out’ → ‘sen-dout’ (linking, /t/ links to /ou/)]
Brianna: He did, but I had already started it and I wanted to get it out of the way.
[‘wanted to’ → ‘wanna’; ‘get it out’ → ‘geddit-out’ (connected, flap T)]
Marcus: That’s smart. Hey, do you want to grab a bite to eat? A bunch of us are going to that Thai place on Fifth.
[‘do you want to’ → ‘djuh wanna’; ‘bite to eat’ → ‘bite-tuh eat’ (linking); ‘bunch of us’ → ‘bunch-of-us’ (fully linked chain)]
Brianna: Yeah, I’d love to! Let me just grab my coat and I’ll meet you at the elevator.
[‘I’d love to’ → ‘I’d love-tuh’; ‘Let me’ → ‘Lemme’ (assimilation); ‘meet you’ → ‘meetcha’ (assimilation)]
Marcus: Sounds good. Tell Linda to come too if she wants. The more the merrier.
[‘Tell Linda’ → ‘Tel-Linda’ (gemination of /l/); ‘more the merrier’ → fully linked phrase]
Brianna: I’ll ask her. She mentioned she had to pick up her kids, but I’ll check.
[‘mentioned she’ → ‘mentiondshe’ (linked); ‘pick up her’ → ‘pick-up-ur’ (full linking chain)]
Marcus: Okay! See you in a minute.
[‘See you’ → ‘Seeya’; ‘in a’ → ‘inna’ (linking + reduction)]
Brianna: Be right there!
[‘Be right’ → natural linking; ‘right there’ → ‘right-there’ (flap T possible for right: ‘ri-there’)]
Notice how nearly every phrase contains at least one instance of connected speech. This is normal. When you listen to natural American English, your goal is to recognize word boundaries even when sounds have shifted. With practice, your ear will begin to decode these patterns automatically.
About the Author
