Understand spoken language

Breakdown of words, phrases and dialogues

Submitted by admin on 5 November 2014

The breakdown of words, phrases and dialogues describes how each of the entries on Lingopolo are broken down into their component parts.

100% match

The basic rule is that the breakdown should match 100% the word, phrase or dialogue.

This means that there is 100% match in the sound: when a user plays the word, phrase or dialogue, and then plays each of the parts in the breakdown they should hear exactly the same thing. Note that hearing exactly the same thing is about the exact number of words and the exact form of those words; the voice and accent of the breakdown can and often does differ.

This also means that there is 100% match in what is written: when a user looks at the word, phrase (or dialogue), and looks at what is written in the breakdown they should see exactly the same thing.

Thus, if the French is "elle a cinq ans", all the words in the breakdown, "elle", "a", "cinq", "ans", when listened to individually, should not be missing any parts in the original, nor should have any parts in the recording not heard in the original, and similarly all the written parts if joined together would equal the original.

Examples:

Note too that even slight differences in orthography, such as a singular or plural form with just an "s" at the end of a word, or a slight variant in verb form, e.g. "parle" vs. "parles" will need a different breakdown. See for example the examples in use for do (2nd person singular) vs. does; is doing (3rd person singular).

Note that you can use the "clone content" button to create an identical copy of a given word as a starting point.

Exceptions

There is a slight exception for the written version; when the phrase is longer than a certain amount, then it may be truncated. For example for a sentence like "the 23 things I like are...", then we can if necessary use a means such as "..." or "etc." to indicate that the original phrase is not shown in full. In general this should be avoided and only used in very long sentences.

For dialogues, the whole dialogue contents are not written in the title, but just a summary, often with the word "dialogue" in brackets, e.g.

Another exception to the 100% sound and written match is when the natural breakdown is different, for example:

  • of the (masculine) here we want to show that "du" is made up of "de" + "le", but the sound and the written forms are different

What degree to breakdown?

Generally the rules are as follows:

  • phrases are broken down into individual words
  • long phrases are broken down into smaller phrases
  • dialogues are broken down into phrases, not into indivdual words
  • compound words are kept together as a single group, and then broken down at the lowest level
  • at the lowest level, stop, don't breakdown an entry into itself

Phrases

Generally phrases are broken down into individual words. This means that every word in the phrase is added individually to the breakdown. For example:

Why is it necessary to break down to such detail? It is important for several reasons.

Firstly, Lingopolo can then automatically create a list of all uses of a particular word in use. For example, look at the "Examples of 'a (masculine)' in use".

Secondly, when phrases are broken up into the smallest possible parts, Lingopolo can automatically calculate the most frequently used words and then uses that as the basis for being able to teach users the most important words first.

Sometimes a phrase is quite long. In that case you will need to break the phrase down, not as individual words, but as sub-phrases. See Long phrases, including a discussion of when is a phrase a long phrase and when is it a normal phrase.

Numbers

Numbers are a general case of compound words. You should always keep a large number together as a single group and then break it down at the lowest level. For example:

Why? Because this way the numbers over 20 lesson will automatically have all the numbers. If the above phrase were broken down as "my" + "older" + "sister" + "is" + "30" + "5" + "years" + "old", then the number 35 does not exist as a valid word.

Dialogues

A dialogue is two or more phrases which belong together.

A dialogue may actually be a monologue (i.e. just one person speaking), or a story, but for Lingopolo purposes we call everything that has two or more phrases a dialogue.

Dialogues are broken down into phrases, not into indivdual words.

(Each of the phrases is of course further broken down into the individual words, as for any phrase, but the dialogue itself is broken down into phrases.)

The dialogue should generally be broken down into phrases which individually make sense, e.g. see breakdown of Didier introduces himself.

Where a dialogue can naturally be split into several sections, then each section should be made a separate dialogue. For example, many of the FSI dialogues are in several parts, e.g. the shopping dialogue is in fact in 3 parts, the man getting a lift to the shopthe man buying some shirts, and the man then visiting the shoe department. We therefore make 3 seperate dialogues. We also create a mega-dialogue which groups together all the parts, e.g. Let's go shopping (dialogue in 3 parts)

Examples:

Wherever possible you should use the audio from the original dialogue to create the phrases in the dialogue breakdown. Do not rerecord the phrases yourself, but rather use Audacity to extract the phrases from the original dialogue recording. This has various advantages:

  • it simply sounds more natural to have the phrases in the dialogue sound like the original in the dialogue
  • the student, who may have trouble understanding the whole dialogue, can hear just an individual phrase from the dialogue 100% as it sounds in the dialogue; same speed, same accent, same background noise, same music, same sound effects...
  • it gives more variety of recording voice in individual phrases and sentences. It is not always in the context of the dialogue that a student may hear and practise a particular phrase. For example a phrase like, I am 36 years old is from the dialogue Muriel introduces herself and is in the voice of Muriel, but the student may hear it in the context of practising the numbers. This variety of different voices is something both nice for the student, but also of teaching value to them to get used to hearing different voices.

Here are a few examples which show how the original sentences have been pulled directly from the original dialogue:

It may happen that a dialogue contains a short sentence, such as "hello" which is already recorded. In that case it is OK to link to the already existing page, even though it is not taken directly from the dialogue being broken down.

The full list of published dialogues is of course at Dictionary of Dialogues.

Slow to begin, but faster and faster

The breakdown of a phrase is slow to begin with, since almost all of the words in the breakdown do not yet exist in Lingopolo, and thus need recording and a seperate page adding. As time goes on though, the breakdowns become easier and easier to do quickly since more and more of the words already exist on the site, so it is just a question of adding them into the breakdown fields.

A work in progress

Language is not always straighforward, and deciding the best way to breakdown a phrase and whether to split a word up or to keep a sub-phrase together as a whole is tricky. I'm still learning the best way to break some of the trickier combinations down, and so as new examples arrive, some slight changes might happen.