Cyberbullying speech patterns among Indonesian studentsLanguage evidence is consulted in dealing with cyberbullying cases. Linguistic analysis is used to reveal the nature of speech and cyberbullying expressions that have legal implications. This study aimed to find the speech patterns used by students in cyberbullying, identify the features of insults in cyberbullying speech, and identify the legal implications of cyberbullying speech on social media. 504 speech data were collected from Facebook and Twitter from January to April 2020.  They were all written in social media accounts of Indonesian high school students. The cyberbullying speech data was analyzed qualitatively using forensic linguistics analysis, i.e., analyzing the linguistic aspects of social media speech data with linguistic theory, particularly speech act theory. Results of the linguistic analysis were then linked to the legal implication of the social media speech. The analysis shows that four general patterns of cyberbullying speech are found, Head Act, Head Act - Supportive Move, Supportive Move - Head Act, and Supportive Move - Head Act - Supportive Move. In addition, three features of insults in cyberbullying are found: the use of pejorative words or phrases, the use of words or phrases with negative connotations, and the use of 'reply' or 'mention' features. The patterns of cyberbullying speech found in this study indicate that the insulting effects of different word or phrase formulation vary. However, they all may have legal implications, especially the violations of Law of Information and Electronic Transaction (UU ITE), particularly Article 27 Paragraph (3) regarding insults and defamation.Keywords: cyberbullying, social media, speech act, insults, forensic linguistics Pola tuturan perundungan siber (cyberbullying) di kalangan pelajar IndonesiaAlat bukti kebahasaan sangat diperlukan dalam penangangan kasus perundungan siber. Pembuktian tuturan perundungan siber dilakukan melalui hasil analisis linguistik yang mampu menunjukkan hakikat pertuturan dan ekspresi perundungan yang berimplikasi hukum. Tujuan penelitian ini adalah menemukan pola tuturan yang digunakan oleh pelajar dalam melakukan perundungan siber, mengidentifikasi fitur ekpresi penghinaan yang terkandung dalam tuturan perundungan siber, dan mengidentifikasi implikasi hukum tuturan pada perundungan siber di media sosial.  504 data tuturan yang berasal Instagram dan Twitter diambil pada rentang waktu Januari hingga April 2020. Seluruh data tuturan tersebut ditulis oleh akun pelajar sekolah menengah di Indonesia. Analisis data kualitatif dilakukan untuk mengkonfirmasi data tuturan perundungan siber dengan telaah linguistik forensik, yakni menganalisis aspek kebahasaan data tuturan media sosial dengan teori linguistik, khususnya teori tindak tutur. Hasil analisis linguistik kemudian dihubungkan dengan aspek implikasi hukum dari tuturan media sosial tersebut. Berdasarkan hasil analisis, ditemukan empat pola umum tuturan perundungan siber, yaitu head act, head act - supportive move, supportive move - head act, dan supportive move - head act - supportive move. Selain itu ditemukan pula tiga fitur ekspresi penghinaan dalam tuturan perundungan siber, yaitu penggunaan kata atau frasa peyoratif, penggunaan kata atau frasa berkonotasi negatif, dan penggunaan fitur membalas (reply/mention). Pola tuturan perundungan siber yang telah ditemukan dalam penelitian ini menunjukkan perbedaan efek penghinaan yang ditimbulkan dari formulasi kata atau frasa, namun semua tuturan tersebut memiliki potensi implikasi hukum, khususnya pelanggaran UU ITE, utamanya pasal 27 ayat (3) tentang penghinaan dan pencemaran nama baik.Kata kunci: perundungan siber, media sosial, tindak tutur, penghinaan, linguistik forensik

This paper will present a small part of the results of a study conducted in 2018, aiming to identify ten graders' ability to read. The study involved 6,539 students from 298 schools (non-PISA) in 34 provinces in Indonesia. This paper will centre around the students' ability to comprehend information presented in multimodal texts. The data were obtained from a reading test consisting of 40 questions, 16 of which involved information gained from multimodal (verbal and visual) texts. The data were analysed using descriptive statistics. The results show that in general students still need guidance to understand information presented in multimodal texts. The average of the correct answers of the 16 questions involving verbal and visual texts was 45.46%. A question asking students to synthesise information from the picture was answered correctly only by 9.24% participants. Moreover, questions to do with the relationship between visual and verbal texts were answered correctly by less than 40% participants. However, questions on information familiar to the students, such as the location of a new book in a library, could be answered correctly by 87.05%. It is recommended that students and teachers should be exposed to multimodal teaching materials and teachers should be trained to deal with multimodality and to explain the relationship between visual and verbal texts explicitly to the students in all subjects.