Most people seem to think that the London Underground was designed to transport people quickly and efficiently around the city, the various lines being linked up one to another at interchange stations to make for ease of connection. But some English teachers - and most buskers - would think otherwise! For anyone who teaches English to foreigners, the London Underground seems to have been purposely designed to teach linking - no, not connecting between one tube line and another, but connecting the end of one word to the beginning of the next. Linking occurs constantly and naturally in normal speech but often causes problems for learners. And so where better to practise it, but on the underground?

Before we actually take a trip or two on the underground, let's get down to mechanics ... the mechanics of linking words together. In English, when we speak at a normal speed - and even more so when we speak quickly - we join our words together. Some foreigners not used to such a way of speaking in their own languages say that we "eat our words" or "swallow our words". In short, we tend to link the end of one word to the beginning of the next, at times even modifying the actual pronunciation in the process.

This linking process is not a haphazard one; there are specific patterns governing the phenomenon. When talking about linking, it is useful to be able to refer to the kinds of sounds heard at the end of one word or the beginning of the next and describe them as either 'consonant sounds' or 'vowel sounds'. By the term 'consonant sound' we mean the first or last sound in a word, not its last letter. In other words, the last sound of Moorgate (to take the name of an underground station) is the consonant sound /t/, even if the last written letter is actually a vowel. Likewise, the final sound in 'high' is a 'vowel sound', despite the spelling. Here is a brief overview of the main forms of linking. (Note that many individual sounds will be shown using the International Phonetic Alphabet, and written between slash marks, as with the /t/ sound above.)

• Consonant-to-Vowel Linking. This occurs when the first word ends in a consonant sound and the next word starts with a vowel sound. In a phrase such as 'large eggs' we would actually hear something sounding more like 'lar-jeggs'; likewise 'put on' would sound more like 'pu-ton'.
• Vowel-to-Vowel Linking. Here both the end of the first word and the start to the second word have vowel sounds. It is useful to make a subdivision into three kinds of final vowel sounds where linking takes place:
1. Words ending in 'ee' sounds (/i:/ or /I/ in the phonetic alphabet) link to the next word with a small 'y' link (known as 'yod' with the IPA symbol /j/). 'Happy Anniversary' therefore sounds something more like 'Happy-yanniversary'.
2. Words ending in 'oo' sounds (/u:/ or /U/) link to the next word with a small /w/ sound. So 'New Orleans' becomes something like 'New-Worleans'.
3. Words such as 'car', 'where' or 'over' that end with a silent 'r' (in Received Pronunciation in England - Americans and Scots usually pronounce such 'r' sounds anyway) often regain a short /r/ sound. For example, 'far off' becomes something closer to 'fa-roff'.
• Consonant-to-Consonant Linking. This is where one word ends and the other word starts with a consonant sound. When a /t/ or a /d/ sound occurs at the end of a word, in many cases it simply disappears before another consonant, though a very short break (what a linguist would call a 'glottal stop', and shown by a /?/ in the IPA) might replace the lost sound. Therefore 'get by' would sound more like 'ge-by'. With a few specific initial consonants, namely /k/, /g/, /p/, /m/ or /b/, a phenomenon known as 'assimilation' takes place.
• Assimilation, as the name suggests, occurs when the end-of-word consonant sound and the following initial consonant sound tend to merge together and become similar. As we say the last sound of the first word, we anticipate the mouth position for the first sound of the new word and the two become mixed. So a final /t/ in 'white gold' will sound more like a /k/ as we anticipate the mouth shape for /g/, and therefore we hear something like 'whike-gold'. The precise patterns for consonant-to-consonant linking and assimilation are as shown in the following table:

Original sound /t/ /t/ /t/ /d/ /d/ /d/ /n/ /n/

Changing almost to (dropped) /k/ /p/ (dropped) /g/ /b/ "ng" /m/

In front of many consonants /k/ /g/ /p/ /m/ /b/ many consonants /k/ /g/ /p/ /m/ /b/ /k/ /g/ /p/ /m/ /b/

Example phrase get back (ge-back) a light glass (a-like-glass) Great Britain (Grape-Britain) mind the gap (mine-the gap) a bad case (a-bag-case) old men (olb-men) a thin girl (a-thing-girl) a clean plate (a-cleam-plate)

All of this sounds very complicated, but fortunately most linking forms should come fairly spontaneously to students if they can be persuaded to take care to join their words together. In fact, students will automatically use linking if they do not pause between words and if they speak at a reasonably quick (normal) speed.

It is therefore important to make students aware of this phenomenon, especially if linking is not common in their own mother tongue - German mother-tongue speakers, for example, need a lot of encouragement to link their words in English, to become a little more 'sloppy' in their pronunciation. Some explanations will probably be useful, but it should all happen automatically in practice, and that's when it becomes time to take a trip on the underground.

Beg (please don't steal) or borrow a map (or maps) of the London Underground - or you may find that maps of other underground systems from the English-speaking world can be just as useful. Show students how many of the names of stations are made up of two parts. Usually these two parts are two separate words 'Barons Court', 'Chalk Farm', 'Marble Arch' and so on; occasionally they are written as one word 'Highgate', 'Westferry' or 'Westminster'. Whether they are written as one word or two, linking may occur.

Look at a few examples of the names and start to identify some patterns that occur. Many two-name stations don't require any of the linking features for a smooth pronunciation. Shepherd's Bush, Charing Cross and Piccadilly Circus - to take just three - easily roll of the tongue as they stand. But a large number of stations provide excellent examples of the various kinds of linking we've already described.

Consonant-to-vowel linking is noticeable in station names such as: Warwick Avenue (Warwi-Kavenue), North Acton (Nor-Thacton), East Acton (Eas-Tacton), Marble Arch (Marbe-Larch) and Mile End (My-Lend), as well as Dagenham East, West India Quay, Burnt Oak, and many more.

There are good examples of vowel-to-vowel linking, too. Highbury and Islington has a 'yod' link (as well as a consonant-to-vowel slur), making it sound something like 'Highbury-yan-Dislington'. There are also a number of examples of /w/ links, such as Harrow and Wealdstone (Harro-wan-Dwealstone), as well as Harrow on the Hill and Hounslow East.

Consonant-to-consonant links abound. Final /t/ and /d/ sounds are dropped in many station names, such as Bond Street (Bon-Street), East Finchley (Eas-Finchley), Old Street (Ol-Street), Westferry (Wes-ferry) and so on. Assimilation is also well demonstrated: Eastcote (Eask-cote), West Brompton (Wesp-Brompton), Westminster (Wesp-minster), Wood Green (Woog-Green), Redbridge (Reb-bridge), Kilburn Park (Kilburm-Park) and Hatton Cross (Hattong-Cross).

All in all, the people who devised the station names for the London Underground have kindly given us over 50 examples of names where linking occurs. And station names are ideal, complete little phrases for illustrating and practising the phenomenon. Moreover, for students learning in Britain or for those likely to be making a future trip to London, it really kills two birds with one stone!

After some brief explanations, you can perhaps let students loose on an underground map to try to find as many examples of linking as they can, or perhaps just a few good examples of one kind of linking at a time so as not to confuse the issue too much. For further practice, students could work for a few minutes together, reading the stations on any particular line. (It is, however, best to avoid using the eastern, outlying end of the Circle Line where some names are particularly difficult to pronounce.)

A bit of general conversation can make use of the station names, too: Who's been to London? Did you stay near an underground station? Which one? Where did you take the tube to? Where did you change? If you want to get from ... to ..., where will you have to change? and so on, checking up on linking all the time.

The London Underground is not the only good place to practice linking, however. There are other areas of language where linking is prominent. Abbreviations provide good examples, particularly of vowel-to-vowel linking: BA, KLM, TWA, VIP, CNN, TNT, etc. Telling the time is another simple area of language where linking is very common, with consonant-to-vowel and vowel-to-vowel linking being thick on the ground (a useful counterpart to the London Underground map where consonant-to-consonant linking and assimilation are the most common). When telling the time, 'twelve o'clock' sounds more like 'twel-voclock', 'three o'clock' becomes something along the lines of 'three-yoclock', 'a quarter to eight' has a /w/ link in 'to-weight', 'half past nine' will drop the 't' in 'pas-nine', and so on.

A game called Time Lines (nothing to do with tenses) is a fun way of practising lots of linking by combining times and names of underground stations. The only equipment needed is a map of the underground (a big wall map or small individual ones). The object of the game is to describe a journey on the underground, station by station, adding five minutes' journey time between stations. At each interchange station, changes of lines are allowed (and even encouraged), not least to make sure that everyone is following. Students can work in groups, or in smaller classes the teacher can say whose goes next. The game should run something like this (explanations of the various steps are given in italics in the brackets):

• Student 1 At four o'clock we were at East Acton. (we're on the Central Line)
• Student 2 At five past four we were at White City. (we're going east)
• Student 3 At ten past four we were at Shepherd's Bush.
• Student 4 At a quarter past four we were at Holland Park. (still on the Central Line.)
• Student 1 At twenty past four we were at Notting Hill Gate. (an interchange station - the next student can decide to change to another line, or continue on the Central Line)
• Student 2 At twenty five past four we were at High Street Kensington. (we've changed to the District Line or the Circle Line)
• Student 3 At half past four we were at Earl's Court. (we're on the District Line, for sure, but this is another interchange station)
• Student 4 At twenty to five we were at South Kensington. (now we've changed to the Piccadilly Line)
• ... and so on.

In these eight short sentences, there are over twenty examples of linking! As it might take a little while before the class gets the hang of the game, it's best to start as a whole class and out in the suburbs before dividing into groups and travelling into the centre where line changes can be very numerous. In fact playing the game in central London becomes rather chaotic, but if it's all in good fun and if students can keep their minds on the linking, then it's an enjoyable way of approaching a rather intricate pronunciation problem. At the end of the game, the class will probably agree with you that connections on the underground are only slightly easier than connections between words!

   Published in Modern English Teacher (Vol. 9/1, January 2000), Pearson Education Ltd, Basingstoke, UK.