Deception criteria

The hypothesis of our empirical research was that verbal signs of falsehood can be identified in written texts. Consequently, 15 markers that are likely to point at the falsity of the written text were identified and tested in different language backgrounds. The data are taken from authentic texts written by students in two languages: Ukrainian (their mother tongue) and English (first foreign language).
Two hundred students voluntarily took part in the research. They were asked to describe events on a given topic (summer vacation) in two texts: one of them holding true information and the other message being false. One hundred participants wrote two texts each in Ukrainian, and one hundred participants (with estimated levels of English as B1-C1) produced truthful and false accounts in English.
Student's t-test and Wilcoxon's T-test were used to categorize a statistically significant difference, as well as Spearman's correlation analysis was used to assess the relationship between the indicators.
The data confirmed that regardless of the language (either Ukrainian or English) texts with true statements tend to be longer than the ones containing false information in most cases. They are also more intensively filled with emotional-sensory and spatial details, contain numerous references to the speaker’s opinion, experience, feelings (self-reference). False statements tend to actualize concepts different from those that are conveyed in true texts written by the same author on the same topic. The comparison of Ukrainian and English texts written by Ukrainian students showed that falsehood markers in the latter embrace ‘abundant’ lexical diversity (type-token ratio), irrelevance and/or redundancy of information, repetitions, and infrequent numerals. Ukrainian written texts containing untrue information, however, demonstrated violations in the chronology of events as well as context changes unjustified by the narrative perspective. Evidence shows that the following features were statistically insignificant, though present in ‘false’ written speech for both languages. These include the proportion between the prologue’s length to the entire text volume; the use of modifying adverbs; the use of ambiguous, non-specific nominations; spontaneous corrections in writing; frequent use of pronouns (the first person plural as well as the third person singular and plural).
Conclusion. Statistical evidence shows that a number of verbal units help distinguish a false written text from a true one, despite the fact that by its nature false speech is a hidden code and imitation of any type of true speech. However, these criteria differ for texts written in the native and foreign languages, namely in Ukrainian and English.

本实证研究之假设为：虚假的口头符号可以在书面文本中识别。因此，我们识别并测试了15个可能指向书面文本虚假性的标记，这些标记在不同的语言背景下进行了检验。数据来源于两位语言背景下的原创文本：乌克兰语（母语）和英语（第一外语）的学生所写。两百名学生自愿参与研究，他们被要求就给定主题（夏季假期）撰写两个文本：一个包含真实信息，另一个为虚假信息。一百名参与者分别用乌克兰语撰写两个文本，而另一百名（英语水平估计为B1-C1）的参与者则用英语撰写真实和虚假的叙述。为了区分具有统计学意义的差异，本研究采用了学生t检验和Wilcoxon T检验，同时使用Spearman相关性分析来评估指标之间的关系。数据证实，无论语言（乌克兰语或英语），含有真实陈述的文本在大多数情况下往往比包含虚假信息的文本更长，且更密集地填充了情感-感官和空间细节，包含了大量对说话者观点、经验、感受的引用（自我参照）。虚假陈述往往实现与同一作者在相同主题上所写真实文本传达的概念不同的概念。乌克兰学生所写乌克兰语和英语文本的比较显示，后者中的虚假标记具有‘丰富’的词汇多样性（类型-标记比）、信息的不相关性和/或冗余、重复以及不常用的数字。然而，包含虚假信息的乌克兰语书面文本却表现出事件时间顺序的违规以及由叙事视角无法解释的语境变化。证据表明，以下特征在两种语言的‘虚假’书面语言中均具有统计学上的不显著性，尽管它们存在于书面语言中。这些包括序言长度与整个文本体积的比例；修饰副词的使用；模糊、不具体的命名；写作中的自发更正；频繁使用代词（第一人称复数以及第三人称单数和复数）。结论。统计证据表明，一些口头单位有助于区分虚假书面文本与真实文本，尽管虚假语言本质上是一种隐藏代码和对任何类型真实语言的模仿。然而，这些标准在母语和外语文本（即乌克兰语和英语）中有所不同。