Quick Reference Guide
This product has been discontinued and is no longer supported.
- What PhoneBox does
- Getting Started
- How to Run PhoneBox
- How to Use the Files Created by PhoneBox
- Improving the Quality of Your Phonetic Data with Your Language Informer
- Preparing a List of Words to Help You Learn To Hear Differences (Hearing List)
- Interpreting Ambiguous Segments and Sequences
- Looking for Contrast
- Importing Charts into Microsoft Word
- Importing Interlinear Text into Microsoft Word
- Other Uses
- Error Handling
What PhoneBox does
PhoneBox is designed to help Toolbox users to do phonological analysis on their data. If you have your phonetic data in a Toolbox database, let’s say a MDF – like dictionary database, you can:
- Generate a table of all phones in your data, and of how often they occur.
- Generate environment charts of all phones in your data.
- Mark phones as uncertain in your data and then create charts of all uncertain phones, giving you all examples where you’re not sure about your phonetic data.
- See how the interpretation of ambiguous segments or ambiguous sequences affects the syllable patterns of your data.
- Create a chart of all segment sequences in your data matching a given cv-pattern.
These lists and charts are created as Toolbox databases, so you can use Toolbox to view, sort and filter them to help you analyse them, and to export them to a word processor for printing. There are also two Microsoft Word document templates which allow you to import PhoneBox phone inventory tables, combination charts and Toolbox interlinearized text into Word for formatting and printing.
The default settings that come with PhoneBox assume that you use the MDF database type for your data and Unicode UTF-8 for phonetic notation. But you don’t have to. With PhoneBox, you can use any database type and any font your data happens to be available in. However, you should consider reencoding your data into Unicode format, if you have not done so already.
PhoneBox has the following requirements:
- Windows 98 SE or later, or WindowsNT 4.0 SP6 or later.
- Microsoft Visual Basic 6 Runtime. Is normally present on Windows systems with these versions. If PhoneBox doesn’t run on your system immediately, you can download the runtime files separately.
- Toolbox. PhoneBox relies on the Toolbox project file and other settings files in order to find necessary information on the character encoding of your data. If you are still using Shoebox: PhoneBox should be able to run with the Shoebox settings files, too, but all the default settings files make use of the Toolbox Unicode features and have to be changed before you can use them.
- If you want to use the Word document templates to create phone charts and combination charts or to work with interlinear text, you need to have Microsoft Word 97 or newer. The document templates included in this version of PhoneBox assume that your data is in Unicode notation. If you need a different format, please contact me.
- Before you install a new version of PhoneBox, please uninstall any old version using the Add / Remove Programs (or Software) control panel in Windows Settings.
In order to install PhoneBox, run
XXXis the version number.
The following sections assume that you are familiar with the Toolbox terminology, i.e. database, database type, language encoding, record, field, marker, sorting, filtering, browse view, template record etc. For reference information on these terms, please see the Toolbox online help or other Toolbox materials.
Preparing Your Project for PhoneBox
In order to work with PhoneBox, you should set up a Toolbox project dedicated to the analysis of the phonology of your language. At the beginning, the project should contain the database with the data you want to analyse. Since PhoneBox needs certain database types and language encodings for the files it creates, these need to be present in the project folder. In Order to prepare your project for PhoneBox, please:
- Choose PhoneBox / Settings Folder from the Start menu. The Settings folder will open.
- Copy all the files (database types, language encodings and the phonebox.bat file) to your project folder, if you don’t already have them.
- You need at least the five PhoneBox database types, but you may modify them as you like, as long as you don’t delete the markers that are defined there. The language encodings are default encodings for use with Unicode phonetic notation end the Doulos SIL font. If they conflict with the encoding of your data, you can replace them with your own language encodings.
- Open your project in Toolbox and create a new database, using the “PhoneBox Settings Unicode” database type.
Creating a PhoneBox Settings Database
PhoneBox needs to know:
- What data to process
- How to interpret the data
- Where to put the results of processing the data
- What to do with the data
All this information has to be presented to PhoneBox in a special database called ‘settings database’. When working with PhoneBox, you will create a settings database, create a record in this database telling PhoneBox what to do (this will be referred to as ‘settings record’), run PhoneBox on this record and then look at the files PhoneBox created by opening them with Toolbox. A settings database has got to use the database type ‘PhoneBox Settings’ which comes with the PhoneBox program. This database type contains a template record you can use to start analysing your data, and documentation on the markers themselves. You can read this documentation by right-clicking on the markers or choosing ‘Database Types…/PhoneBox Settings/Modify…/Markers/Modify…’ from the Project menu.
The following list gives short information on the different sections in a settings record and on all the markers of the “PhoneBox Settings” database type. More information can be found in Geoffrey Hunt’s PhoneBox Tutorial.
Since you can use any font for your phonetic data, you’ve got to tell PhoneBox which characters in your font mean what in terms of phonological analysis. Each character you use in your data has to be put into exactly one of the following categories: consonants, vowels, length-marks, stress-marks, tone-marks (different groups for diacritics (accents) and spacing tone letters), modification marks (different groups for spacing marks in front of the base character, diacritics, and spacing marks following the base character), ignore characters, which are present in the data but don’t carry any meaning for the analysis, vowels that may be consonants in some environments, and consonants that may be vowels in some environments. In the settings database, there is one field for each of these character groups. You put a character into one of these groups by typing it into the appropriate field, separated from its neighbours by space characters. If you need to use multigraph characters, you just type them in without separating spaces.
The following table summarises all the Character Declaration markers:
- \c : All characters that are consonants. Different segments have to be separated from each other by space characters or line breaks. Multigraph segments are defined by placing the character combinations together without separating whitespace, like “c d f ch ts”
- \v : All characters that are vowels. Different segments have to be separated by space characters or line breaks. Multigraph segments are defined by placing the character combinations together without separating whitespace, like “a e i au ai”
- \dfmf : Definition of Marks in Front of the base character. Must contain all the characters that modify the articulation of the following segment.
- \dfma : Definition of Marks Above or below the base character. Must contain all the diacritical marks that modify the articulation of the base segment. Diacritical marks must be entered after the base segment.
- \dfms : Definition of Spacing Marks. Must contain all the marks that modify the articulation of the preceding segment and take up a character space of their own.
- \t : All the diacritical marks that signify Tone (accents).
- \ts : All the Spacing Tone letters.
- \l : All the characters signifying Length.
- \s : All the characters signifying Stress.
- \ign : All the IGNore characters. These characters might occur in the data, but don’t have any meaning for PhoneBox. They normally don’t appear in phone inventories, and they don’t split multigraph characters if they occur in the middle of them.
- \mbc : May Be Consonant. All the characters that are vowels but may be consonants in certain contexts. This field helps with interpreting ambiguous segments. A MBC remains a vowel in the following contexts:
- Isolated between word boundaries.
- Word initial before a consonant.
- Word final after a consonant.
- Between two consonants.
- Between a consonant and another mbc or mbv, causing the second ambiguous segment to be interpreted as a consonant.
- In all other contexts, it will become a consonant.
- \mbv : May Be Vowel. All the characters that are consonants but may be vowels in certain contexts. This field helps with interpreting ambiguous segments. A MBV turns into a vowel in the following contexts:
- Word initial before a consonant.
- Word final after a consonant.
- Between two consonants.
- Between a consonant and another “May Be Vowel”, causing the second MBV to turn out as a consonant.
- In all other contexts it remains a consonant.
- \syc : Syllable pattern Consonant. The character PhoneBox should use to mark a consonant in a phone inventory or environment chart. “c” by default.
- \syv : Syllable pattern Vonsonant. The character PhoneBox should use to mark a vowel in a phone inventory or environment chart. “v” by default.
All these fields have the “phonetic” language encoding and are displayed in the IPA font. Every character used in your data, except the space character, must be declared in one of these fields. If PhoneBox finds a character in your data that it doesn’t find in one of these fields, it will create an error message. Please see the FAQs on how to deal with this kind of error message.
Another group of fields in the settings database is dedicated to tell PhoneBox where to find your phonetic data. The database with your data is referred to as ‘input database’. This database has got to be a Toolbox database, but it may use any database type. In the settings database, you’ve got to enter the filename of the input database, its record marker, the marker(s) that contain the phonetic data, and a list of markers you want to copy into the environment chart files to help you analyse your data, i.e. part of speech or gloss fields from your input database. Other fields are used to tell PhoneBox about an optional uncertainty status field in the input database. PhoneBox can use the status field to determine what data items are fit to be included in an analysis.
The following table lists all markers that are used to declare what data to process:
- \in : The filename of the input database. Must be relative to the directory which contains the settings database.
- \utt : The markers that contain Utterances, data in phonetic notation. If there are more than one fields in a record that can contain phonetic data, the markers have to be listed with separating spaces, like “lx ph”.
- \cpy : Copy markers. The content of these markers from the input database will be included into the environment charts, to help with analysing the data. List multiple markers with separating spaces, like “ps ge”. If you want a copy of the original utterance, include the utterance markers here, too.
- \fil : Filter. Tells PhoneBox to include only a certain group of base characters in the output. List the base characters like in the \c and \v fields. Optional. If this field is not present or empty, all segments from the input database will be processed.
- \usf : Uncertainty Status Field. Marker of the field that contains the uncertainty status of a given segment.
- In the input database, an uncertainty status field must follow its corresponding utterance field immediately. For each spacing character in the utterance field, there must be either a 1 or a 0 in its status field. A 1 means that the segment has been checked and is certain, a 0 means that the segment has not been checked and is not certain. If there is an empty status field, the whole utterance is uncertain, if there is no status field, the whole utterance is certain. Only data with completely certain segments is included into environment charts and phone inventories.
PhoneBox can produce three different kinds of output files:
- Statistics files (database type PhoneBox Statistics)
- Environment Charts (database type PhoneBox Chart)
- Combination Charts (database type PhoneBox Combinations)
A statistics file contains a list of all different segments PhoneBox found in your data, a count of how often it found each one in your data, and a sign if it interpreted them as consonants or vowels.
The following markers control the creation of statistics files. If the marker is present and contains a valid file name, a statistics file of the requested type is created, otherwise it is not. All file names are relative to the directory which contains the settings database.
- \wt : Phone Inventory of all phonetically certain segments.
- \uwt : Phone Inventory of all phonetically uncertain segments. If you specify anything in this field, you must fill in the \usf field as well.
- \cvwt : Inventory of all segment combinations that match a certain cv-pattern (see combination charts, you must fill in the \cvp and cvxy fields as well.)
An environment chart normally lists all occurences of a given segment in a tabular view, using one column for the segment itself, one for the part of the utterance before the segment, and one for the part of the utterance after the segment. In PhoneBox, an environment chart file works much the same way, with the addition that you get a few extra informations on the utterance: a ‘canonical’ version of the utterance without ignore characters and with tone and stress characters in the right places, a guess of the cv-structure of the utterance, and whatever information from the input database you want to have in the chart.
The following fields control the creation of environment charts:
- \ch : Chart file. The filename for the environment chart of all certain segments in your input database.
- \uch : Uncertain Chart file. The filename for the environment chart file for all uncertain segments in your data. You can use these lists as hearing lists, to recheck your data with a language informer. If you fill in this field, you must fill in the \usf field as well.
A combination chart normally is a table with one set of segments on the x-axis, another set of segments on the y-axis, and examples for existing combinations of the segments in each cell. With PhoneBox, a combination chart looks very much like an environment chart, except that there are two additional fields, which contain the x-part and the y-part of each example combination.
The following fields control the creation of combination charts:
- \cvp : A CV-Pattern that tells PhoneBox which segment combinations to include. The following characters are allowed in this field:
- c: a single consonant. Multigraph characters that are defined in the \c field are counted as one.
- v: a single vowel. Multigraph characters that are defined in the \c field are counted as one.
- #: a word boundary.
- ?: any single consonant or vowel.
- 0..9: any consonant or vowel from a group defined in the fields \cv0 to \cv9, respectively.
- \cvxy : The position that separates the X-part of the CV-pattern from the Y-part. If your cv-pattern in the \cvp field is “cv” and your \cvxy field is “2”, then all the consonants will be in the the \x field and all the vowels that follow them in the \y field of the combination chart.
- \cvch : CV combination Chart. The filename of the combination chart. If you fill this field, you have to fill the \cvp and \cvxy fields as well.
- \cv0..\cv9 : CV-variables 0 to 9. You can list any group of vowels or consonants in one of these fields, like in the \c and \v fields, and then access the variable by using the digits 0 to 9 in the \cvp field to match segments from the corresponding group.
Like with the settings database type, there is specific information on each marker in these databases in the database types which come with PhoneBox.
The following markers are defined in the PhoneBox Settings database type, too:
- \ref : Reference. The record marker of the settings database. This field must be present in every settings record, and it must contain a unique, single word name for the record. When running PhoneBox, you will have to provide the filename of the settings database and the name of a settings record, which refers to this field.
- \rem : Remarks. A short text describing the purpose of the settings record. This field is for your own information, to help you document what steps you took during your analysis.
- \outp : Output folder Path. A folder where to store all the files PhoneBox creates. Optional. Default is the folder which contains the settings database.
How to Run PhoneBox
To run PhoneBox, follow these steps:
- Create a record in the settings database. Fill in at least the ref, in, and utt fields, and at least one of the output filename fields. If you’re using Unicode for your data, you can leave the remaining fields of the template as they are for a first run.
- Choose ‘Run Batch File…’ from the Tools menu in Toolbox.
- Type “phonebox” and press the “run” button.
- PhoneBox will start and show a wizard dialog, asking you to specify the project, the settings database and a record from this database. Once you pressed the “Finish” button, PhoneBox will close Toolbox and start to process your data.
- When finished, PhoneBox will display a message window to tell you the outcome of the processing. If all went well, there’ll be just a message telling you how many records have been processed. Otherwise, PhoneBox will display an error message. Please refer to Error Handling on what to do then.
- PhoneBox will cause Toolbox to open the project again, now including all the files that were created by PhoneBox. The files are displayed in a default view, preconfigured for the most common tasks for each file type.
How to Use the Files Created by PhoneBox
The following sections tell you very briefly what you can do with the files you generated. More information is available in the German Daba Tutorial, and will hopefully soon be available in English, too.
Improving the Quality of Your Phonetic Data with Your Language Informer
To do this, you have to enter uncertainty status information into your input database. Do the following:
- Create an environment chart of the uncertain data (fill in the uch and usf markers of the settings record).
- Look at the uncertain environment chart. Select Browse View and choose sgm, syl and utt (or the original utterance marker) as browse fields.
- Put a filter on the sgm field returning only the segments you want to investigate. Normally, that would be a group of similar segments. Sort by syl, bef, aft. This will group the list into groups with similar syllable structure and the uncertain segment in similar positions. Normally, this will help your language informer to tell you whether there is a difference between two similar segments or not.
- Print the list, or export it to Word via RTF Export for printing. Then you can let the language informer decide which occurences of the segments are equal and which differ, and make a hypothesis on what might be the difference.
Preparing a List of Words to Help You Learn To Hear Differences (Hearing List)
Do the same as above with the normal environment chart file, only using the ‘Syllable Analysis’ filter from the PhoneBox Chart database type.
Interpreting Ambiguous Segments and Sequences
To analyse syllable patterns and to test interpretation of ambiguous segment and sequences, set up your project like this:
- Generate a Phone Inventory to show the segments in your data before interpretation.
- Generate an environment chart and load it into Shoebox.
- Look at the environment chart. Switch to Browse View and choose the following browse fields: utt, syl
- Activate the predefined “Syllable Analysis” filter from the PhoneBox Chart database type.
- Sort the environment chart using the syl and the utt field (in this order) and the ‘Syllables’ sort order (should be default).
- Scrolling up and down the list, take stock of the syllable patterns you can guess at from the syl column. Identify suspicious cv-patterns that might be simplifyied by reinterpreting ambigous segments.
- Move suspicious vowels from the \v field to the \mbc field in the settings database, suspicious consonants from the \c field to the \mbv field.
- Rerun PhoneBox and look if you’ve eliminated the suspicious cv-patterns. Caution: If you put a segment into the mbc field, PhoneBox tries to guess whether it should be a vowel or a consonant on each occurence. The guessing algorithm will interpret a single word initial i (for instance) followed by another vowel as a consonant, and will leave an i between two consonants alone, but if the situation is more complicated, PhoneBox might guess wrong, so look at your data carefully to check if you find the guess plausible… Repeat until you’re satisfied (or frustrated 😉 ).
- Look for further suspicious cv-patterns that could be eliminated by grouping ambiguous sequences into single segments.
- Enter the sequences into the c or v fields of the settings record. You must enter modification characters if they occur in the middle of a multigraph sequence, but not if they precede or follow it.
- Rerun PhoneBox and see if you’ve eliminated some syllable patterns. If yes, you’ve got evidence that your sequences might be segments of their own.
- After you have reached a satisfactory interpretation of ambiguous segments and sequences, copy the input database and retranscribe it using search and replace, using unmistakable notations that make the interpretation clear. Adapt your settings record to the new notation, too.
- Generate a new Phone Inventory to document your interpretation.
Looking for Contrast
To look for contrast, set up your project in the following way:
- Generate a statistics file.
- Look at the statistics file and identify a group of segments you want to investigate
- Create a new record in the settings database. Fill in the \ch field to create an environment chart, and the \fil field to process only the segments from the group you want to look at.
- Look at the environment chart. Add fields from the cpy marker to the Browse Fields as you see fit. Some field which tells you whether two different entries have the same meaning would be helpful.
Looking at your data this way, you will have similar or equal environments next to each other, and you can look at how your segments are distributed and if there’s contrast. Especially, minimal pairs should be right above each other, and using a filter with the ‘Non-Unique’ filter element will return all minimal pairs immediately.
Importing Charts into Microsoft Word
To import a Consonant Chart, Vowel Chart or Combination Chart into Word and create a formatted chart, do the following:
- Generate a Statistics file or Combination file with PhoneBox.
- Choose “Phonology Charts” from the PhoneBox group in the Start menu. If Microsoft Word is installed, this will cause it to create a new document based on the Phonology template. If Word asks whether to activate macros, please choose Yes.
- Follow the instructions in the template to import the Statistics file or Combination file.
The template assumes that your data is in Unicode format. If you need it in a different encoding, please contact me.
Importing Interlinear Text into Microsoft Word
To import interlinear text into Word, do the following:
- Interlinearise text with Toolbox, following the instructions in the Toolbox online help.
- Choose “Interlinear Text Table” from the PhoneBox group in the Start menu. If Microsoft Word is installed, this will cause it to create a new document based on the Text Table template. If Word asks whether to activate macros, please choose Yes.
- Follow the instructions in the template to import the interlinear text into a table in Word and to rearrange it until it meets your needs.
One could think of other uses, like filtering on the syl field to investigate the influence of the syllable structure on tone and stress. If you found another possible way to look for a phonological phenomenon, or want to do a certain analysis on your data and don’t know if it is possible with PhoneBox, please contact me, I’ll be happy about any feedback.
In earlier releases of PhoneBox, the error messages were a constant source of misunderstandings and inconvenience to the users. In PhoneBox 2.0.0, it was a major goal to improve the error reporting features of PhoneBox. This release features a new message window that is aware of the encoding of the data and of the font used to display it, so that it will display characters from your data the same way they look in Toolbox.
If anything goes wrong, the reasons might be in three different categories:
- Something is wrong with the settings.
- Something is wrong with the input data.
- Something is wrong with PhoneBox itself or the way it is installed.
Errors in the Settings
If an error in the settings prevents PhoneBox from processing your data, PhoneBox tries to give a specific error message with advice on how to correct it. In that case, PhoneBox will not create any files. You should then change the settings record you were working with according to the hint and try again.
Errors in the Data
If anything is wrong with your input data, the message will refer to a database called “errors.db”. In this case, PhoneBox will process all records from your input database that it can use, and ignore the ones that cause problems. When Toolbox opens again, You will find a database called errors.db in your Toolbox project. By browsing through it, and using the Toolbox parallel jump feature, you will find all the records in the input database that could not be processed, and a short message about what is wrong with them. (Credits to Karen Buseman for suggesting this approach).
System errors are caused by programming mistakes or problems with the installation of PhoneBox. In case of a system error, PhoneBox will also display an error message window. The error message will mention an “errLogfile.log” file in the PhoneBox program folder. This file is probably not very helpful for the user, but it will help me to find out what went wrong. So, if you keep receiving system errors, please report them to me, and include the log file with your report.
If you encounter a system error, or don’t find out what to do with a certain error message, you can save the error message to a RTF file using the “Save…” button on the message window. Please put the error message, all your project files and, in case of a system error, the errLogfile.log, into a zip file and mail it to the support email address. If you are not using Unicode or SIL IPA 93 for phonetic notation, please include also the font you use to display your data.