ETSI TR 101 568 V1.1.1 (2012-02)

Human Factors (HF); A study of user context dependent multilingual communications for interactive applications

DTR/HF-00137

General Information

• Status: Published
• Publication Date: 02-Feb-2012
• Technical Committee: HF - Human Factors
• Current Stage: 12 - Completion
• Due Date: 01-Feb-2012
• Completion Date: 03-Feb-2012

Standards Content (Sample)

Technical Report
Human Factors (HF);
A study of user context dependent multilingual
communications for interactive applications

2 ETSI TR 101 568 V1.1.1 (2012-02)

Reference
DTR/HF-00137
Keywords
application, HF, ICT, interaction, language, user
ETSI
650 Route des Lucioles
F-06921 Sophia Antipolis Cedex - FRANCE

Tel.: [phone]  Fax: [phone]

Siret N° 348 623 562 00017 - NAF 742 C
Association à but non lucratif enregistrée à la
Sous-Préfecture de Grasse (06) N° 7803/88

Important notice
Individual copies of the present document can be downloaded from:
http://www.etsi.org
The present document may be made available in more than one electronic version or in print. In any case of existing or
perceived difference in contents between such versions, the reference version is the Portable Document Format (PDF).
In case of dispute, the reference shall be the printing on ETSI printers of the PDF version kept on a specific network drive
within ETSI Secretariat.
Users of the present document should be aware that the document may be subject to revision or change of status.
Information on the current status of this and other ETSI documents is available at
http://portal.etsi.org/tb/status/status.asp
If you find errors in the present document, please send your comment to one of the following services:
http://portal.etsi.org/chaircor/ETSI_support.asp
Copyright Notification
No part may be reproduced except as authorized by written permission.
The copyright and the foregoing restriction extend to reproduction in all media.

© European Telecommunications Standards Institute 2012.
All rights reserved.
TM TM TM
DECT , PLUGTESTS , UMTS and the ETSI logo are Trade Marks of ETSI registered for the benefit of its Members.
TM
3GPP and LTE™ are Trade Marks of ETSI registered for the benefit of its Members and
of the 3GPP Organizational Partners.
GSM® and the GSM logo are Trade Marks registered and owned by the GSM Association.
ETSI
3 ETSI TR 101 568 V1.1.1 (2012-02)
Contents
Intellectual Property Rights . 5
Foreword . 5
Introduction . 5
1 Scope . 7
2 References . 8
2.1 Normative references . 8
2.2 Informative references . 8
3 Definitions and abbreviations . 9
3.1 Definitions . 9
3.2 Abbreviations . 11
4 Localising an interactive application . 11
4.1 Localisation vs. translation . 11
4.2 Localisation aspects . 12
4.2.1 Localisation and language complexity . 12
4.2.2 Grammatical aspects . 12
4.2.3 Social aspects . 14
4.2.4 Cultural aspects . 14
4.3 Localisation of context dependent applications: games . 15
4.3.1 Flavour of the game localisation problem . 15
4.3.2 Variable and modifier: Example of interactivity script . 18
4.4 Localisation of all context dependent applications . 20
4.4.1 Sectors sensitive to context-dependent applications localisation . 21
4.4.2 Example of interactivity script: 1-to-1 real-estate sector . 22
4.5 Impact on interactive applications . 23
4.5.1 Localisation process . 23
4.5.2 Localisation impact on the economical world . 24
4.5.3 Localisation impact on the society . 24
5 Localisation sensitive sectors . 25
5.1 Context dependent application industry . 25
5.1.1 Differences in industry size not sector . 25
5.1.2 Technical limitations of existing solutions . 25
5.1.3 Existing technologies for interactive application localisation . 26
5.1.4 Existing proprietary solutions for interactive application localisation . 27
5.2 Limitations of traditional ICT localisation rules . 31
5.2.1 Optimized source texts . 31
5.2.2 ICT localisation guidelines . 32
5.2.3 Missing localisation aspects . 33
6 Language technologies' state-of-art . 35
6.1 Machine translation (MT) . 35
6.1.1 General . 35
6.1.2 MT technologies . 35
6.1.3 Advanced terminology management technologies . 36
6.2 Multilingual dialogue systems . 36
6.2.1 General . 36
6.2.2 Human-computer dialogue . 37
6.2.3 Multiparty dialogue. 37
6.2.4 Current limitation of Multilingual Dialogue Systems . 37
7 Generic analysis . 38
7.1 Localisation requirements of the industry . 38
7.1.1 Necessity for absolute correctness for all languages . 38
7.1.2 Key localisation success factors . 39
ETSI
4 ETSI TR 101 568 V1.1.1 (2012-02)
7.2 Localisation environment architecture requirements . 40
7.2.1 Localisation environment principles . 40
7.2.2 Localisation result for the application end-users . 40
7.2.3 Localisation environment for the application developer . 41
7.2.4 Localisation environment for the localisers/translators . 42
7.2.5 Localisation environment for the society . 43
8 Conclusions and recommendations . 44
Annex A: Linguistic complexity . 46
A.1 Grammatical number . 46
A.2 Grammatical gender . 47
A.3 Grammatical cases . 48
A.4 Other grammatical specificities . 49
Annex B: Bibliography . 51
History . 52

ETSI
5 ETSI TR 101 568 V1.1.1 (2012-02)
Intellectual Property Rights
IPRs essential or potentially essential to the present document may have been declared to ETSI. The information
pertaining to these essential IPRs, if any, is publicly available for ETSI members and non-members, and can be found
in ETSI SR 000 314: "Intellectual Property Rights (IPRs); Essential, or potentially Essential, IPRs notified to ETSI in
respect of ETSI standards", which is available from the ETSI Secretariat. Latest updates are available on the ETSI Web
server (http://ipr.etsi.org).
Pursuant to the ETSI IPR Policy, no investigation, including IPR searches, has been carried out by ETSI. No guarantee
can be given as to the existence of other IPRs not referenced in ETSI SR 000 314 (or the updates on the ETSI Web
server) which are, or may be, or may become, essential to the present document.
Foreword
This Technical Report (TR) has been produced by ETSI Technical Committee Human Factors (HF).
Introduction
Of the study objectives
The present study addresses the issue of localisation for multilingual context-dependent interactive applications.
"Localisation" is the process of adapting the application for a specific country. It implies not only translation of
dialogues or phrases from one language to another, but also the adaptation of idiomatic and cultural characteristics. The
same issue is relevant to multilingual interactive applications where several languages are to be supported
simultaneously.
The objective of the study is to define how to simplify the development process of highly interactive multilingual
applications and to ensure the top quality of their localisation.
The present study is intended for anyone dealing with complex localisation of context-dependent interactive products,
such as dynamic online systems, video games, serious game and eLearning, smartphone applications, internet-based
applications accessed by PC or mobile, etc. It concerns applications designers, developers, publishers, product managers
and distributors, as well as all stakeholders who may benefit from its use, including service and application providers,
end-users, etc.
In short, the present study primarily concerns those who have already experienced a serious localisation problem, when
designing or using an application, especially designers, publishers, users or all those who don't want to experience this
problem at all.
Therefore, the present study tries to achieve two key goals:
• To describe the state of art in the localisation issues and techniques, especially regarding context dependency;
• To describe a way forward, a proposed roadmap leading to guidelines or potentially standard in that area.
Although this roadmap may need additional collateral information, the study shows that the contents, the scope and the
potential solutions for such guidelines is clearly defined, so that the Technical Committee may launch these guidelines
study with no delay, nor additional research required.
Of the study background
ICT users are becoming increasingly involved and fully immersed in applications such as video games or one-to-one
Internet-based applications. The more immersed the user, the more successful the application! Two key factors
determine the extent of such immersion: an increasingly realistic environment (such as graphics), and a more in-depth
textual or oral interaction. Applications therefore demand "online" textual or oral interactivity with the user in a
complex, accurate and natural-sounding way. Texts are created on the basis of the user context, which, in turn, depends
directly on the user's actions and his/her environment.
ETSI
6 ETSI TR 101 568 V1.1.1 (2012-02)
The complexity of dialogues and interactions with the environment in different context has become so important that it
is practically impossible to plan for every potential combination. Text "strings" (chains of characters) to be created by
the application are therefore constructed dynamically from scripts for human-machine dialogues, through "engines"
generating at real-time phrases that are dependent on these context variables.
Once created, these applications are to be adapted into languages or countries different than the original ones they have
been created for- a process known as "localisation". It implies not only the linguistic translation of dialogues or phrases
from one language to another, but also the adaptation of idiomatic and cultural characteristics. In simple applications,
with little dynamically generated text, the localisation process includes the translation of the whole User Interface (UI)
and text strings from the source language into the target language.
This is however not possible for interactive applications based on variables and interactivity scripts. Localising such an
application implies translating all possible UI and text strings from one language into another, identifying all variables
and their potential values, and also translating these variable values into the target language. The fact that majority of
applications are being written in English or Asian languages, which have a very simple grammar system, increases the
difficulties when translating into other, more complex-structured languages, where grammatical agreements vary
depending on case, number and gender. The problem becomes even more critical when having to adapt cultural
variables. These issues can lead to limiting the number of countries in which the application can be marketed.
Alternatives are either forcing users to use English, or releasing poor quality applications in localised languages, risking
a poor audience or worse, a negative buzz.
Several types of industries are facing this critical problem, such as the game industry, education, telecom, internet,
automotive industry, etc. Many of them are working around the problem by simplifying the dialogues to avoid
grammatical barriers, therefore reducing the quality and the level of immersion. And there are no emerging languages
technologies able to propose a valuable solution yet.
Therefore, there is a strong need, both for the designers and for the end users, to study issues relevant to the localisation
of such context dependent multilingual interactive applications, approached from all relevant stakeholders' perspectives,
to understand the complexity and specificity of the issue throughout all the involved application segments, to analyse
how these sectors are addressing or working around the problem, and finally how the whole application development
community can define together a common way to solve this increasingly critical issue.
Of the study boundaries
The study will focus on text-based interactivity, since this is the core of all communications, even audio ones. Indeed,
applications are either explicitly text based (messages are displayed to the user or taken from him through keyboard) or
they add an audio interface, as input or output. Audio inputs are based on Speech To Text (STT) and Automatic Speech
Recognition (ASR) technologies to be able to record and process user input. Audio outputs are either pre-recorded
audio (then static and with no link with our scope) or based on Text-To-Speech (TTS) technologies able to generate
speech out of a dynamic text. The present study will then not consider audio at all, and the speech technologies STT,
TTS, ASR, although of high interest in multi-language systems, will not be presented in the report as being totally out
of scope.
It is also important to explain what part of the concept of UGC – User Generated Content – the present study is
covering. UGC is a generic term covering all types of information, used in a broad range of applications, which is
coming directly from the user, such as news, forums, comments, blogging, digital video or images, podcasting, etc. In
the present study, since target applications are context dependent interaction, UGC is restricted to the user context,
including his profile, his inputs, his history and previous actions, etc., which could be recorded in real-time context
variables, as in 1-to-1 marketing, or games or role playing in eLearning scenarios. Typically, user input such as
comment, chat, discussion, is out of scope. However, user input asked for a name, an answer to a question, a decision, a
choice, are to be considered. In a first phase, "closed" inputs only will be taken into account, and later more open and
informal answers.
ETSI
7 ETSI TR 101 568 V1.1.1 (2012-02)
1 Scope
The present document gives an introduction to and an analysis of the most important issues and areas of relevance to
context dependent multilingual communications for interactive applications. It provides a clear description of the most
common difficulties and problems faced by application designers and localisers today, and how they solve or work
around these.
The scope of the present document is summarised through the following statements about the study:
1) It defines localisation and explains what is involved in the localisation process of interactive application,
including the management of interactive and non-interactive applications translations.
2) It describes in detail problems and issues related to the localisation of interactive applications, to help
understand the limitations, needs and existing solutions or work-around used in the field.
3) It identifies the different industrial and technical domains that are directly concerned by the issue. It looks at
several innovation activities related to the domain and provides a state-of-art presentation of languages
technologies and research in the domain of multilingual applications, translation management and localisation.
4) It collects information from the different industry sectors identified, examining their needs, their localisation
process, their management of translations, and potentially, specific tools or processes they are using for
solving or working around the problem.
5) It provides a generic analysis of the situation, plus a specific analysis related to each identified industry sector.
The present document addresses the localisation process, which covers a large spectrum of issues and activities.
However, the study, after providing an exhaustive description of what localisation means, will focus on the localisation
aspects relevant to the highly interactive applications heavily using context variables.
The present document had an initial focus on the game industry but it does not restrict its scope to video games. On the
contrary, it will expand its vision beyond games, aiming at identifying all other technical and economical sectors facing
similar issues.
LOCALISATION COMPLEXITY
Social
and Cultural
SCOPE
Context Variables
and high interaction
Grammar
correctness
Basics
LOCALISATION
SENSITIVE
INDUSTRY
Figure 1: Technical Report Scope diagram
Finally, the present document will not provide an exhaustive analysis of advanced research projects and techniques, but
it will review the main existing or known areas of innovation and analyse whether they may help solving our problem.
ETSI
8 ETSI TR 101 568 V1.1.1 (2012-02)
2 References
References are either specific (identified by date of publication and/or edition number or version number) or
non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the
referenced document (including any amendments) applies.
Referenced documents which are not found to be publicly available in the expected location might be found at
http://docbox.etsi.org/Reference.
NOTE: While any hyperlinks included in this clause were valid at the time of publication ETSI cannot guarantee
their long term validity.
2.1 Normative references
Not applicable.
2.2 Informative references
The following referenced documents are not necessary for the application of the present document but they assist the
user with regard to a particular subject area.
[i.1] ETSI EG 202 417: "Human Factors (HF); User education guidelines for mobile terminals and
services".
[i.2] W.L. Johnson, S. Marsella, N. Mote, H. Vilhj´almsson, S. Narayanan, and S. Choi: "Tactical
language training system: Supporting the rapid acquisition of foreign language and cultural skills".
In Proceedings of InSTIL/ICALL2004 - NLP and Speech Technologies in Advanced Language
Learning Systems, 2004.
[i.3] W.L. Johnson, S. Marsella, and H. Vilhj´almsson: "The DARWARS Tactical Language Training
System". In Proceedings of the Interservice/Industry Training, Simulation, and Education
Conference (I/ITSEC), 2004.
[i.4] R. Klischewski: "No man's tool - Why we need games localization tools". LISA China
Conference. Suzhou, China, 2010.
[i.5] US patent "System and method for generating grammatically correct text strings" - US patent
office - number 7983895, July 19, 2011.
ETSI
9 ETSI TR 101 568 V1.1.1 (2012-02)
3 Definitions and abbreviations
3.1 Definitions
For the purposes of the present document, the following terms and definitions apply:
avatar: See "playing character".
crowd sourcing: it is the act of outsourcing tasks, traditionally performed internally within a company or a group, to an
undefined, large group of people or community
NOTE: It can be a community design (all or part of a development task), a community review and evaluation, or
any other community activity.
context variables: See "variables".
dynamic dialogue (vs. static dialogue): user-application dialogue where interactivity and sentences depends on the
user context, the user actions and his environment, therefore with an infinite number of possible phrases
end user: See "user".
function: abstract concept of a particular piece of functionality in a device or service
ICT devices and services: devices or services for processing information and/or supporting communication, which has
an interface to communicate with a user
localisation: advanced process, which consists of adapting an application for a specific country, not only through the
translation of dialogues or phrases from one language to another, but also through the adaptation of idiomatic and
cultural characteristics
manual: See "user guide".
Machine Translation (MT): automatic translation, as by computer, from one natural language to another
NOTE: Initially restricted to word to word translation, current MT systems are using technologies such as
rule-based translation, translation memories, dictionary based techniques or statistical techniques, as well
as hybrid systems using several of these techniques.
modifier or variable modifier: modifier is tag to be put on a variable to indicate that the value of the variable needs to
be modified to be in grammatical accordance with gender, numeral, case, etc.
NOTE: Modifiers could be use for social and cultural aspects also, as, for instance, for indicating formal versus
informal addressing.
non-playing character (NPC): any fictional character of a game or an application not controlled by a player. In video
games, the NPC will usually be controlled by the computer
NOTE: In other games, or in many applications (such as Serious Game), the NPC will be a character controlled
by the game master, the educator or the Master of Ceremony.
1-to-1 marketing/eCommerce: personalized marketing/eCommerce as an alternative to mass marketing. The system
analyses each client or prospect in order to adapt the communication and sales accordingly
player: denomination of any user in a game or a serious game
playing character: any fictional character of a game or an application directly controlled by a player
NOTE: It is also called the "avatar", as the game entity representing the user, its behaviour and acting on his
behalf.
ETSI
10 ETSI TR 101 568 V1.1.1 (2012-02)
sim-ship: simultaneous shipment of a given application in several countries to avoid the need for any additional
investment.
NOTE: Localisation should then be done for each country.
static dialogue (vs. dynamic dialogue): user-application dialogue where interactivity and sentences are well defined in
a fixed tree of user actions or environment, therefore with limited number of possible phrases
Terminology Management System (TMS): translation memory tools (see TM below) use either segment based
translation memory engine or a corpus based engine, or both.
NOTE: This is called a "terminology management system". This ensures consistency in the translation by
automatically searching for previous examples of where phrases were translated in the document.
translation: process of taking textual or oral communication elements, in the form of sentences or phrases, from one
source language and translating them into a target language
Translation Memories (TM): translation memory allows for re-use of what has been translated previously by a
professional translator.
NOTE: There are TM by domains, industry, projects or groups of projects. The importance of terminology is
huge
user: person who uses an application - see also "player"
user education: any information provided to users of a product or service on the functionality provided by the product
or service and any instructions on how this functionality is to be used
User Generated Content (UGC): generic term covering all types of information, used in a broad range of applications,
which is coming directly from the user, such as news, forums, comments, blogging, digital video or images, podcasting,
etc.
NOTE: In the case of the present study, since target applications are context dependent interaction, UGC is
restricted to the user context, including his profile, his inputs, his history and previous actions, etc., which
could be recorded in real-time context variables, as in 1-to-1 marketing, or games or role playing in
eLearning scenarios.
user guide: technical communication documents, intended to give assistance to users using a particular product
User Interface (UI): physical and logical interface through which a user communicates with a telecommunications
terminal or via a terminal to a telecommunications service
NOTE: Also called man-machine interface, MMI.
variable: variable is an element of a sentence that can take a different value depending on the user context and the
game context
NOTE: In highly interactive applications such as game, the possible values of a variable can be quite high, the
combination of variables values being then infinite.
XLIFF: interchange format used in the localisation field, as an extension of XML
NOTE: The format is used widely but does not handle key elements such as variables table, needed for context-
dependent applications.
ETSI
11 ETSI TR 101 568 V1.1.1 (2012-02)
3.2 Abbreviations
For the purposes of the present document, the following abbreviations apply:
ASR Automatic Speech Recognition
CAT Computer Aided Translation tool
CM Content Management
CMS Content Management System
CRM Customer Relation Management
GCMS Global Content Management System
GMS Globalization Management System
HTML Hyper Text Mark-up Language
ICT Information and Communication Technologies
MMORPG Massive Multiuser On-line Role Playing Games
MT Machine Translation
NPC Non-Playing Character
PC Personal Computer
PDF Portable Document Format
R&D Research and Development
STT Speech to Text
TM Translation Memory
TMS Terminology Management Systems
TTS Text to Speech
UGC User Generated Content
UI User Interface
XLIFF XML Localisation Interchange File Format
XML eXtensible Mark-up Language
4 Localising an interactive application
4.1 Localisation vs. translation
As defined, "translation" is the process of taking textual or oral communication elements, in the form of sentences or
phrases, from one source language and translating them into a target language. Translation can be done from a well
defined and fixed text (official translation of an official document or a book, for instance); translation can be also done
simultaneously by human interpreters, while a person is talking in a conference. Translation of applications on the other
hand consists in taking both fixed and dynamically generated texts displayed by the application from one language to
another.
Existing and emerging languages technologies can help to translate applications using semi-automated systems such as
Translation Memory or fully-automated systems such as Machine Translation (Goggle translation, internet automatic
forum or consumers' comments translations, etc). These systems give variable results depending on the available
corpus, the context, the content, the type of text and the format. For instance, Translation Memory, which has now
reached maturity, is often integrated in GMS – Globalization Management Systems or GCMS – Global Content
Management Systems. In the same way, Machine Translation is still limited but it gives great results when following a
strict process (e.g. controlled English during authoring). This is the case of the automotive industry, for instance, which
performs most of its translation using MT through a well-defined user interaction process.
As defined, "localisation" is a more advanced process, which consists of adapting an application for a specific country,
not only through the translation of dialogues or phrases from one language to another, but also through the adaptation of
idiomatic and cultural characteristics.
The document EG 202 417 [i.1] provides a good and clear definition of the localisation process and its implications for
mobile terminals and services, which is fully applicable to the context dependent interactive applications:
"Localisation refers to the provision of product and user-guide variants for different markets taking into account local
linguistic and cultural differences. This presents a special challenge as the ICT market is a global market and most
manufacturers try to market their products globally. In many countries, the localisation of consumer products is required
by regulation.
ETSI
12 ETSI TR 101 568 V1.1.1 (2012-02)
As the costs for localising products and services are considerable, most manufacturers and service providers restrict
their localisation efforts to offering different language versions of the user interface (in particular in the menus) and of
the user guides. The use of icon-based menus (currently state of the art at least on the main menu level) is an attempt to
internationalise aspects of the user interface of many applications.
One of the main challenges related to localising applications is that as localisation is done fairly late in the development
process, the localisation efforts of application comes after completion of master draft, and even sometimes after the first
release in other countries, so always at a very late stage. Since all last-minute changes to the master also have to be
made to all language variants, correct and complete language variants are only available in later editions".
The following clauses will explain the differences between translating and localising and therefore the technical and
cultural aspects of localisation. Although the localisation of context dependent interactive applications may have some
common aspects with other ICT sectors, such as mobile terminals and services, there provide a lot of highly specific
aspects that make us understand why solving the problem is difficult and, as today, not solved at all.
4.2 Localisation aspects
4.2.1 Localisation and language complexity
The grammatical complexity of languages is endless, and the examples provided in this clause are only giving a quick
flavour of problem. A more in-depth presentation of the most common complexity elements of different languages is
provided in Annex A. Cultural and social aspects are adding to this "technical" complexity a human and emotional layer
that is even different, within the same language, between cultures.
This aspect could have stay a minor and useless question in the way translation and localisation was several decades
ago: why bother when your application is speaking English, German, French, Japanese and Chinese and when sentences
to be translated are the text bars in menus or error messages? But, as explained in the introduction, three major
phenomena are now forcing industries to address the problem of localisation fast:
• Applications should now offer a totally immersed interaction in an increasingly realistic environment. Users
are not accepting to be forced in to a foreign language (English) or gender (male) any more.
• Globalization enforces distributors to provide a given application simultaneously in an average of
12 languages per release (typical industry standard), and to add additional language for each target markets.
• Minority policies and markets are pushing distributors to more and more provide applications in minority
languages, either for social and economical reason (access to eSociety) or for cultural reasons (strong
mobilisation, dense market, cultural heritage preservation).
As a result, localisation is becoming a critical aspect of commercial growth, which increases the pressure onto
applications providers for porting applications onto a broad range of languages, whatever their complexity and cultural
specificities. This is why understanding these complexities and solving the related localisation problem are the 2 steps
that are described in the present study.
Here are several basic elements showing the complexity of languages and why localisation can become uneasy.
4.2.2 Grammatical aspects
• The basic grammatical specificities in languages such as gender, number, cases, plurals: e.g. French has four
types of definite article for the English "the" ("le", "la", "l'" or "les"), and adjectives agree with gender and
number; Russian has two plural forms Finnish has 14 different groups into which you can sort almost every
noun and adjective, etc. So a simple sentence in English for instance, may have many possible translation
depending on the context:
"$Player$ picks up the %color %item"  � Alan ramasse le couteau bleu
� Alan ramasse la perle noire
� Alan ramasse l’épée bleue
� Alan ramasse les pièces rouges
� Alan ramasse la blanche colombe….

Some languages even have different plural depending on the actual numerals. Here is an example from English to
Polish localisation of a very simple phrase:
ETSI
13 ETSI TR 101 568 V1.1.1 (2012-02)
%numeral red flag(s)" � 1 red flag  or  x red flags

Figure 2: Example of Polish localisation of numerals
• Some languages are handling plural accordings, in very specific ways. Some put just an S at the end (most EU
languages) or at least another variant of the noun, while some languages just repeat the noun twice.
e.g.: the man / l’homme (sing.) � the men / les hommes (pl)  in English and French
orang - a person" (sing) � orang-orang – the people " in Indonesian
• Some languages differentiate between the collective, which is indifferent in respect to number, and a set of
single entities (called "singulatives"). For example, in Welsh, moch ("pigs" as a whole) is a basic form,
whereas a suffix is added to form mochyn ("pig" as a single one).
• In other languages, "singulatives" can be regularly formed from collective nouns:
e.g.: Standard Arabic:  حجر ḥajar "stone" → حجرة ḥajara "(individual) stone",
بقر baqar "cattle" → بقرة baqara "(single) cow".
• Numerals may be handled not just as singular or plural in some languages. For instance, in Russian, there are 3
different types of according for numeral: Nominative singular when 1 unique object, genitive singular for 2, 3
or 4 of the object, and genitive plural for any number above 4.
• Possessive pronouns in German agree to both the gender of the owner and of the object, while in French
accordance is only to the gender of the object:
e.g.: Son frère � you don't know whether the 'owner' is male or female
• The dialect variants of particular languages: e.g. Dutch in the Netherlands and in Belgium (Flemish), German
in Switzerland, Austria and in Germany; French in French Canadian and Creole, etc. This includes spelling-
only variants, such as the OUR vs. OR between British and American English (e.g. colour vs. color).
• Translation may vary depending on the context. Some words have several meanings, not the same from a
language to another, so that localisation then takes context and situation into account.
e.g.: (English) "I’ve got the POWER" � (French) "J’ai le POUVOIR" ou "J’ai de la PUISSANCE"
• Additional idiomatic words can be added to a sentence, which have no translation, but add an inflection,
information the speaker want to give: an advice, an order, a question, a doubt, etc.
(French) Vous devriez ouvrir la porte de gauche.
(English) You should open the door on the left first, shouldn’t you?
• Some languages, such as German for instance, are creating very long words and sentences for expressing a
specific action or situation. A simple phrase in English or Chinese, with a certain length, will be translated into
a German phrase more than twice as long. It is not uncommon for short texts, such as the titles of text
commands and menu items, to be three times as long in German as they are in English, while the Chinese
equivalent will be much shorter. For example, the English word "Redo" translated to German is
"Wiederherstellen" up from 4 characters to 16, i.e. representing an expansion of 400 % !
• That might be (actually is) a problem when the application has been designed with strings with a specific
maximum number of characters. Localisation should then adapt the sentence to fit in the maximum string
length, often by using abbreviations, hoping that the user will understand them:
ETSI
14 ETSI TR 101 568 V1.1.1 (2012-02)

Figure 3: Example of text length limitation in German
At some point, even for a German speaker, the abbreviations may be so numerous that it is impossible to understand
what they are finally referring to.
4.2.3 Social aspects
• The use of formal addressing: in some cultures, it is appropriate to address the user using formal language
("Vous", "Sie", "U", etc.), while in others an informal addressing ("Tu", "Du", "Jij", etc.) may be expected
(e.g. French, Spanish or Italian) while other languages would expect the informal form (e.g. German or
Swedish). And, in some cases, dialogues may expect an informal addressing with a formal response.
• Beyond the formal addressing, the functionality of the application itself can be based on hierarchy relations,
such as chief to soldier in a game, or the manager to his employees in an eLearning management training. This
hierarchy is expressed not only through formal addressing, but also through specific cultural coding: e.g. the
same order or advice will be given in a very different tone and words in German (straight, strict) or in Japan
(indirect, soft).
• Interactivity takes into account the social behaviour and social relationship between playing or non-playing
characters. Localisation should then use the profile of the users/actors of the dialogues (static information) as
well as the context, the situation, the intention (dynamic information). This will lead to vary between formal
vs. polite, familiar vs. casual, friendly vs. unfriendly, informative vs. directive, handled different from one
language to another, etc.
• Some languages use a different word depending on who is speaking, and its relation with who he speaks to: "I
show you" has four possible translations depending on gender of the speaker and the listener (i.e. male to male,
male to female, female to male and female to female). Other languages (Thai or Camb
...