Much of how SPSS reads a file in a language is dependent on which version of Windows you are using. Windows 2000 and XP have strong support for displaying characters that are different from the native operating system. Other versions of Windows have varying support.
How any version of Windows reads characters is based on code pages. Code pages are a compilation of characters that are specific to a language or region. Files that use a language from the same code set as you are using can be read without any special considerations. For example, SPSS on an English version of Windows can read files from a Swedish operating system. The code pages are as follows:
Microsoft Windows Codepage : 1252 (Latin 1) - Western standard
Microsoft Windows Codepage : 1250 (Latin 2) - Eastern Europe
Microsoft Windows Codepage : 1251 (Cyrillic)
Microsoft Windows Codepage : 1253 (Greek)
Microsoft Windows Codepage : 1254 (Latin 5)
Microsoft Windows Codepage : 1255 (Hebrew)
Microsoft Windows Codepage : 1256 (Arabic)
Microsoft Windows Codepage : 1257 (Baltic)
Microsoft Windows Codepage : 1258 (Vietnamese)
Microsoft Windows Codepage : 874 (Thai)
Double byte code pages are:
Japanese
Korean
Traditional Chinese (Taiwan)
Simplified Chinese (PRC)
If the file you are reading in has a language from a different code page, it may still be possible to use the file in SPSS. With Windows 2000 and XP, you can change your system locale to the language the file is in. The system locale is located in the Regional Settings Control Panel. If the language you wish to use is not available from the drop down list, then scroll through the Language Settings for System dialog box, and check the locale you would wish to use. Click Apply, and then OK. Reboot if prompted. The language should now be available under the drop down list. Clicking on one locale may add many new languages to your drop down list. You may need a Windows CD in order to install support for the locale.
Once the new system locale has been selected, you may need to run a command like this from within SPSS:
SET LOCALE=Japanese.
(Use SHOW LOCALE to confirm that you got the locale you expected.)
Please note: SET LOCALE command can change from one character encoding to another. A sav file must be valid in the current SPSS LOCALE setting. That is, characters in variable names must be legal in the current SPSS locale, and text strings, while not validated, may not work correctly in character operations if the character set is wrong. In multi-byte encodings, an incorrect LOCALE setting can cause SPSS to become unstable.
If you do not wish to change your system locale, or you do not have Windows 2000 or later, often selecting a font that contains the foreign language characters of the file will allow the file to be displayed properly. However, you may not be able to edit those fields. Depending on your operating system, these fonts may already be installed on your system, or you may need to obtain them from a different source.
If you have the fonts for the language, they can be selected under View->Fonts in the Data Editor. This will allow the data, variable and value labels to be displayed in the character set of the language. You will also need to specify that font in the Output Viewer for the characters to appear appropriately. Most operating systems do not have full right to left script support, so Arabic and Hebrew characters may appear backwards. Be sure to set the "script" field in the font dialog correctly.
Changing the font may not prevent variable names from being changed, due to what SPSS may see as invalid characters. In that case, changing the system locale would be necessary.
