Copyright©2018 P-ISSN: 1978-8118 E-ISSN: 2460-710X 389 Lingua Cultura, 12(4), November 2018, 389-394 DOI: 10.21512/lc.v12i4.4322 THE FORM OF ARGUMENTS USING ARGUMENT RECONSTRUCTION IN ONLINE MEDIA ARTICLES Hestiyani Parai1; Endry Boeriswati2; Miftahulkhaira Anwar3 1,2,3Language Education, Postgraduate Program, Universitas Negeri Jakarta Jl. Rawamangun Muka Rt. 11/ Rw. 14, Rawamangun, Jakarta 13220, Indonesia 1hesti.parai@gmail.com; 2endry.boeriswati@unj.ac.id; 3hera_unj@yahoo.co.id Received: 22nd January 2018 /Revised: 21st May 2018 /Accepted: 11th July 2018 How to Cite: Parai, H., Boeriswati, E., & Anwar, M. (2018). The form of arguments using argument reconstruction in online media articles. Lingua Cultura, 12(4), 389-394. https://doi.org/10.21512/lc.v12i4.4322 ABSTRACT This research aimed to identify the form of arguments in online media articles using argument reconstruction. The data were argumentative texts written in online media article in 2017 in www.hipwee.com, malesbanget.com, and www.idntimes.com. The data studied were 44 paragraphs consisting of 150 sentences. The approach used was qualitative approach by using content analysis method. This research used the triangulation technique (combination). The simultaneous data collection was coupled with data credibility tests with various data collection techniques and sources. Meanwhile, means of documentation and content analysis techniques were done to collect the data. The results indicate that the often used reasonings in online media articles are the generalization in 11 paragraphs and categorical syllogism in 28 paragraphs. The results of this research can be used to find out and develop ways of writing online media articles for writers and readers of online media articles. Keywords: arguments, argument reconstruction, online media articles INTRODUCTION An argument cannot be separated from the use of language. It is because language is a communication tool to convey a message from communicator to communicant. As the argument proceeds, it is certain that the communicator will process the language to convey their message and give an opinion. It is impossible for the communicator to use a language that is not understood by the communicant. If it happens, the message or opinion will not be accepted by the communicant. In addition, people who want to argue must be good at choosing words so that the others can be affected. Argumentation is a specific type of discourse that is often used in daily life (Mochales & Ieven, 2009). The study of argumentation is crucial in many areas of artificial intelligence and text processing research such as reasoning agents or discourse analysis. Moreover, argumentation is a form of rhetoric that seeks to influence the attitudes and opinions of others, so they believe and act according to what is desired by the speaker (Keraf, 2008). Argumentation is also the process of making arguments intended to justify beliefs, attitudes, and values so it can affect others. The primary requirement to write an argument is that the writer must be skilled in reasoning and developing a logical idea (Warnick et al. in Ridhani, 2013). Argumentation can also be a statement both in the forms of oral and written aiming to influence and convince the listener and reader with accompanying evidence of the truth of the opinion to agree with what is said by the speaker or the writer. The primary purpose of writing an argumentation is to convince the reader and to accept or assume certain doctrines, attitudes, and behaviors (Finoza in Tusriawati, Saman, & Martono, 2014). Based on these explanations, an argument is a process for objective data, facts, or evidence so that the truth can be accepted. Its activity includes identifying assumptions to conclusions as it is not much different from logic. It can be concluded that logic is the science of arguments and the argument itself is logic. The practical argument is associated with a set of critical thinking abilities. Critical thinking involves the ability to examine problems, questions, or situations, to incorporate all available information about what is said, to make a solution or hypothesis, and to justify one’s position. The writer can point out evidence to substantiate the reasons given (Ardianto, 2015). Moreover, reasoning is the activity of regularly thinking by using a certain pattern of thinking (Suriasumantri, 2015). Therefore, thinking is not always logical such as daydreaming. Daydreaming is a thinking activity that does not have a specific pattern. It means that someone who lies is doing a thinking activity. This is because the thing the person 390 LINGUA CULTURA, Vol. 12 No. 4, November 2018, 389-394 does is the act done consciously and requires reasoning so the lie can be believed. Good reasoning is necessary to expose ideas in discourse especially in scientific discourse (Brown et al. in Eriyanti, 2017). Disclosure of ideas or topics by using good reasoning in the discourse is at various levels. Those are sentences, fragments, and discourses that have characteristics in accordance with the capacity of each level. Judging from its form, the argument is differentiated into two types. First, deductive reasoning is divided into five types. There are the generalization, structural analogy, metaphorical analogy, lending analogy, and causal relationship. Second, inductive reasoning is divided into three types. There are a categorical syllogism, hypothesis syllogism, and alternative syllogism. In conducting an assessment of an argument, the researchers use the argument reconstruction method. Implicit premises play a huge role in argumentative discourse (Bigi & Morasso, 2012). Furthermore, they have also said that implicit premises are crucial aspects of arguments. Therefore, the argumentation analyst has to be aware of their presence and could reconstruct them. The task of reconstruction does not only distinguish implicit and explicit information, but it also determines whether the language is an argument, and decides what statements are the premises and conclusions. Four elements that enable the researchers to reconstruct arguments systematically are a perspective on argumentation, two discourse principles, an analysis of the rhetorical situation, and mode-specific theories (Van Den Hoven & Yang, 2013). Reconstruction has the same step. The first thing a reader intents on reconstructing arguments is examining the excerpt to determine whether it contains an argument. On some occasions, terms that typically represent inferences allow this to be done. Terms such as ‘thus’, ‘therefore’, and ‘accordingly’ can alert the readers to the likely presence of inference and argument. If the syntactical cues are present, they are likely to be used. After determining that an argument is present in a text, the researchers need to determine what exactly the argument is. Some process is needed to do this has already been accomplished. However, there is also an unfinished process. The next step is to find the missing premise if the conclusion is not written clearly (Katzav & Reed, 2008). Reconstruction has the following steps. Those are identifying discourse (whether there are arguments in a paragraph); identifying the type of argument; identifying premise/conclusion reconstruction; and reconstructing the explicitly represented statements/identifying meaning (Cummings, 1999). As explained earlier, the argument is a process of thinking, and the outcome of this thought process can affect a person. For example, it is online media articles in www.hipwee.com, malesbanget.com, and www. idntimes.com. An argument does not only influence, but it is also capable of changing one’s thinking patterns. For example, there is an article entitled How to Handle Life, For Those Who Have Came Into Adulthood. The article does not include the theoretical sources or survey results of a study related to the title. It may not affect readers who do not read the article. Almost all readers agree with the contents. This suggests that an argument, even without a clear source, it can affect the reader. Based on this, an argument is related to persuasion. In articles writing, arguments expressed by the writer contain persuasive tone to influence the reader in believing the article. METHODS This research incorporates a qualitative research with content analysis method. The review in this research is expected to reveal the pattern of argumentation in online media articles. In this research, argument reconstruction is a tool to analyze data. The data of this research are 44 paragraphs consisting of 150 sentences from www. hipwee.com, malesbanget.com, and www.idntimes.com. The researchers choose these three websites based on the top four rating that they earn in the top 10 ranking sites. Hipwee, malesbanget.com, and idntimes have earned the second until fourth ratings. Moreover, the analysis is conducted from June 2017 to September 2017. There are several steps in analyzing the data. First, there is data collection which includes collecting research object, observing the data, grouping the data in each paragraph and sentence based on the discourse in the form of argumentation, explanation, narrative, and others. Second, data reduction includes data classification. Third, data presentation consists of determining the premise or conclusion that has been listed or unlisted in the data. Fourth, the conclusion is the presentation to the conclusion of the answer to the formulation and a sub-focus of research problems. It is in the form of arguments in the online media articles. RESULTS AND DISCUSSIONS The data consists of some paragraphs. The paragraph consists of several sentences that all components are formed with a specific purpose. To be a good paragraph, the writer requires good reasoning in compiling it. This research explains the form of arguments used by the writers of online media articles using the reconstruction of the argument. The data found in research online media articles published from June 2017 - September 2017 are shown in Table 1. Table 1 Types of Reasoning Types of Reasoning Total (paragraph) % Inductive 16 34% Deductive 28 56% Total 44 100% (Source: Mundiri, 2016) Based on Table 1, it can be seen that from the 50 paragraphs studied. There are 16 paragraphs with inductive reasoning. Then, about 44 paragraphs are deductive reasoning. About 6 paragraphs are not argument paragraphs. Meanwhile, from the perspective of argument reconstruction, the data found are in Table 2. From Table 2, it can be seen that from 150 sentences analyzed. There are 16 sentences which have complete argument reconstruction (ID, IA, IP, IM). However, 134 sentences do not have the argument reconstruction stage. About 109 sentences only have ID and IA, and 25 sentences are up to the third stage of reconstruction argument (ID, IA, and IP). Then, there are several reasonings represented in Table 1. It is inductive reasoning. Inductive reasoning is a procedure that stems from a particular event as a result of 391The Form of Arguments... (Hestiyani Parai et al.) observation. It may also be an opinion that refers to a general conclusion or new knowledge. Based on the results of data analysis performed, there are 16 paragraphs in inductive reasoning. The results are in Table 3. Table 2 Argument Reconstruction Argument Reconstruction Complete Total Less Total ID IA IP IM ID IA IP IM √ √ √ √ 16 √ √ 109 √ √ √ 25 Total 16 134 Total 150 (Source: Cummings, 1999) Description: ID: Identify Discourse IA: Identify Argument IP: Identify Premise IM: Identify Meaning Table 3 Types of Inductive Reasoning Types of Inductive Reasoning Total % Generalization 12 73% Structural Analogy 0 0 Metaphorical Analogy 0 0 Lending Analogy 0 0 Causal Relation 4 27% Total 16 100% (Source: Mundiri, 2016) Generalization is a process of reasoning from some individual phenomena to a general conclusion that binds all similar phenomena to the individual phenomenon under investigation (Irving in Mundiri, 2016). Individual phenomena as the basis of inductive reasoning in this discussion are in the form of propositions (statements). There are 12 paragraphs found. The types of generalizations found in this research are many. Firstly, it is an inductive jump. An inductive jump comes from several facts. However, the facts used do not reflect the entire phenomenon (Keraf, 2008). The data found with this inductive jump is seven paragraphs. For example, there is an article entitled Pro-Kontra Soal Perempuan yang Berpendidikan Tinggi. Bagaimana Pendapatmu? (Pros and Cons of Highly Educated Women. What Do You Think?) in paragraph: Jadi, buat kamu para perempuan yang ingin bersekolah tinggi, sekolahlah! Kalau ada obrolan- obrolan tak enak didengar, yang perlu kamu lakukan adalah menutup rapat dua telingamu. Buktikan kalau keputusanmu tidaklah salah. Tapi ingatlah, pendidikan tinggimu bukan berarti membuatmu bisa berbuat semena-mena terhadap sesamamu. Perempuan terdidik tidak seperti itu. Jadilah perempuan yang berdikari, tapi tetap menaati kodrat yang dimiliki. (So, for women who want to go to high school, go! If there are unpleasant words, all you have to do is close your ears. You have to prove that your decision is not wrong. However, you have to remember that with higher education, it does not mean that you can treat others indiscriminately. Educated women are not like that. Be independent women, but still obey the natural role she has.) This paragraph is said to be a generalization. It is because the last sentence in this paragraph is the main sentence, which is, “Be independent women, but still obey the natural role she has.” This sentence is the main sentence because it summarizes all sentences in the paragraph, from the first sentence about the school to giving advice not to listen to those who discriminate and ask women to remain natural. Seeing the facts in the paragraph, the first sentence emphasizes the reader to continue reading. However, this sentence has not represented the whole issue in the paragraph. The argument reconstruction in this paragraph is its identification of the meaning. There is ambiguity in the second sentence that is, “If there are unpleasant words, all you have to do is close your ears tightly.” This sentence gives a way to ignore those who discriminate by closing the ears. Closing the ears not only means completely covering the ears but also ignoring the words of those people. Second, it is without inductive jumps. Generalization contains sufficient and convincing facts, so there is no chance to argue (Keraf, 2008). The data found with the type of generalization reasoning is in five paragraphs. For example, it is in the article entitled Jenis ‘Virus’ yang Kini Mulai Menggerogoti Generasi Muda Indonesia. Bikin Prihatin Asli! (The ‘Virus’ Type that is Now Beginning to Weaken Indonesia’s Young Generation. So Pathetic!) in paragraph: Jadikan hidup penuh dengan pengorbanan. Semakin menjadi hartawan, semakin pula bertambah dermawan. Semakin terkenal, maka ia pun semakin menjadi teladan. Semakin tinggi suatu jabatan, semakin kebermanfaatan dan kemaslahatan yang selalu dipikirkan. (Make life full of sacrifices. The richer you are, the more generous you become. The more famous you are, the more you become an example. The higher position you have, the more usefulness and welfare you can provide.) It begins with a special sentence because this sentence has not given the main idea of this paragraph yet. The main sentence in this paragraph is found in the last sentence of the paragraph. It sums up the whole paragraph from the sacrifice and generosity. The last sentence unites the four sentences into, “The higher position you have, the more usefulness and welfare you can provide”. The argument reconstruction in this paragraph is there must be a sacrifice in life. The richer a person is, the more worthy he is. Thus, whoever the person is, he must make a sacrifice. The richer he is, or the higher position he has, he needs more sacrifices. However, it does not mean that those who are not wealthy or famous do not need to sacrifice. It is different for each person. Secondly, the causal relationship is reasoning obtained through interconnected phenomena or a dependent relationship between two or more things (Keraf, 2008). It means a result will occur if there is the cause. The causal relationship can take place in three patterns, namely cause to effect, effect to cause, and effect to effect. The data is found on this analogy in four paragraphs. 392 LINGUA CULTURA, Vol. 12 No. 4, November 2018, 389-394 However, the researchers only find two types of causation relationships. The first type is cause to effect. This relation is from an event which is considered as a known cause. Then, it moves forward to a conclusion as the nearest effect (Keraf, 2008). The reasoning of causal relationships with the main sentence is the result in the three paragraphs. The example is in the article entitled Jenis ‘Virus’ yang Kini Mulai Menggerogoti Generasi Muda Indonesia. Bikin Prihatin Asli! (The ‘Virus’ Type that is Now Beginning to Weaken Indonesia’s Young Generation. So Pathetic!) in the paragraph, as follows: Apalagi, pasar-pasar tradisional pun mulai ditinggalkan berganti pasar modern. Seolah tak ada lagi kesempatan untuk tawar menawar secara langsung, bertemu dengan beragam orang tuk saling mengenal, menyapa bahkan mendoakan satu sama lain. Yang semua itu sejatinya semakin membuat kedekatan hati dan kerekatan persahabatan serta persaudaraan antar sesama. (Moreover, traditional markets begin to be abandoned to the modern market. It is as if there were no more opportunities for direct bargaining, meeting various people to know each other, greeting and praying for one another. All of that actually makes the heart closer and strengthens the friendship and the companionship with others.) The data on the cause to effect reasoning is found in the last sentence of, “All of that actually makes the heart closer and strengthens the friendship and the companionship with others.” This paragraph begins with a cause sentence. It is shown by mentioning things like, “Traditional markets begin to be abandoned to the modern market.” The cause sentence supports the next sentence. Indicated by the conjunction of ‘as if’, it shows the similarities that connect two sentences. Therefore, this paragraph is an inductive paragraph with a cause-effect relationship. It starts with a cause sentence. The cause in the sentence is shown in, “It is as if there were no more opportunities for direct bargaining, meeting various people to know each other, greeting and praying for one another.” Then, the general sentence is as the effect of the paragraph which is, “All of that actually makes the heart closer and strengthens the friendship and the companionship with others.” It is an effect because this sentence concludes the paragraph marked by ‘all of that’ which refers to the cause as mentioned earlier. The argument reconstruction in this paragraph is that now the traditional market has changed to the modern market, so there is no longer the intimacy among people. Therefore, today’s changes do not only give a positive impact but also a negative impact such as the relationship between humans. It becomes tenuous due to the little interaction. The second type is effect to cause. This causal relationship is a process from an event considered as the result, then moves to the causes (Keraf, 2008). The main sentence in the causal reasoning of this relationship is found in a paragraph. It can be seen in the article entitled Unggah Status “Marthabak Telor”, Pria Ini Diringkus Polisi (Uploading the Status of “Egg Martabak”, this man was arrested by the police). The paragraph can be seen as follows. Jangan pernah asal mengunggah status di media sosial jika tak mau bernasib sama seperti pria asal Mamuju, Sulawesi Barat ini. Akibat membagikan lelucon di akun Facebook miliknya, pria berinisial H ini pun harus berurusan dengan polisi. (Do not ever upload status on social media carelessly if you do not want to suffer the same fate as this man from Mamuju, West Sulawesi. As a result of sharing jokes on his Facebook account, the man with the initial H must deal with the police.) This paragraph is an inductive paragraph in which the main sentence is a result. Then, the explanatory sentence is the reason. Thus, the type of this paragraph is a causal relationship. The first sentence of this paragraph is the explanatory sentence containing the result. It begins with an explanation of why people should not follow the man who became the actor in this article. It can be seen in this sentence “if you do not want to suffer the same fate as this man from Mamuju, West Sulawesi.” Meanwhile, the second sentence provides support to the first sentence in which the result is proved in the sentence “the man with the initial H must deal with the police.” Based on these findings, the most commonly found data is the generalization and causal reasoning. Generalization is the most commonly used conclusion method by online article writers. This is because generalization is easy reasoning compared to analogical reasoning. For example, in the analogy, there are some things to be compared whereas in generalization anything can be directly created in paragraphs. In relation to online media articles, the writer writes a hot topic among young people. It can be seen from the titles used by the writers. They always use interesting language. Presentation of the article is written in an easy-to-understand language. Based on the goals of each website, it focuses more on their target (young people and Generation Z). Like in the article entitled Jenis ‘Virus’ yang Kini Mulai Menggerogoti Generasi Muda Indonesia. Bikin Prihatin Asli! (The ‘Virus’ Type that is Now Beginning to Weaken Indnesia’s Young Generation. So Pathetic!), it can be seen that the title is very interesting by using the word of ‘virus’. Then, it raises curiosity for the reader. In one of the articles, there is a sentence, “In some of the younger generation, when they are asked what their dreams are, they directly want to be an artist, singer, and others who can be in the tv and have high salary.” It can be seen that the writer understand the desire of young people since the writer mentions things which young people want. Thus, the readers feel that the articles understand what they want. In contrast to generalization, the causal relationship is not used very often by online media article writers. This is because, in the argument reconstruction, the causal relationship does not always have a premise that fills the whole paragraph. In contrast to the generalization with 11 paragraphs, several paragraphs have a complete paragraph construction (causal relationship). Next, there is deductive reasoning. It is a process that generates conclusions based on common ground statements. This reasoning does not require data and facts to support the statement. However, it needs a statement that can identify things correctly so that the taken conclusions are probably true. In this research, the researchers examine some deductive reasoning that is the categorical syllogism, hypothesis syllogism, and alternative syllogism. The percentage of data in deductive reasoning is shown in Table 4. In the categorical syllogism, it is a syllogism consisting of three propositions. Two propositions are 393The Form of Arguments... (Hestiyani Parai et al.) general and specific premises. Then, one proposition is a conclusion. Table 4 Types of Deductive Reasoning Types of Deductive Reasoning Total % Categorial Syllogism 28 100% Hypothesis Syllogism 0 0 Alternative Syllogism 0 0 Total 28 100% (Sources: Mundiri, 2016) The data discovery for this categorical syllogism is 28 data. The example of this syllogism can be seen in the article entitled 6 Hal yang Dilakukan Orang Supaya Dikira Liburan Terus, Kamu Termasuk Gak? (6 Things that would Make People Think that You Always on Vacation, How About You?). The paragraph can be seen as follows. 5. Pamer cerita liburan. Bukan cuma dunia maya, di dunia nyata kamu juga mesti bertindak. Kalau lagi kumpul sama temen-temen, mulailah cerita tentang keseruan liburanmu. Jawablah pertanyaan- pertanyaan mereka dengan antusias, kamu bakal dijadikan panutan mereka deh. Traveler Goals, gitu! (5. Showing off your holiday stories. Not only in the virtual world, in the real world you also have to act. When you get together with your friends, start the story about the excitement of your holiday. Answer their questions enthusiastically, you will be their role model. Traveler Goals!) Showing off holiday stories is the fifth way the writer recommends. After posting holiday photos on social media, the writer assures the reader that they must act real as well as stated in the first sentence (main sentence). Meanwhile, the second sentence explains the steps that can be followed by the reader in recounting one’s vacation in real life. Then, the third sentence continues the explanation of the previous sentence. Thus, the second and the third sentence are special explanatory sentences. The reconstruction in this paragraph is necessary for identifying the meaning and conclusions. There are several meanings that the reader will understand less as it is different from some things that have been mentioned by the writer. In this sentence, the writer tells the readers to tell their vacation experience to their friends. The holiday in question is real, not a made-up one. However, a holiday experience will usually appeal to those who listen to it. The conclusion in this paragraph still needs to be rewritten by the researchers. It tells that the experience of vacation is a way to tell the others that the reader goes on vacation. However, it should be a real experience because in telling a story, there will be questions from listeners who will inquire about the story. As for the other two syllogisms, the hypothesis syllogism is an argument whose major premise is the hypothetical proposition. Then, the minor premise is a categorical proposition that defines or denies antecedent and consequential term to its major premise (Mundiri, 2016). In addition, the alternative syllogism is a syllogism whose major premise is a disjunctive decision. Meanwhile, the minor premise of a categorical decision recognizes or denies one of the alternatives or the major premise (Mundiri, 2016). No data are found for these two syllogisms. This is because the categorical syllogism is a syllogism whose content is in the form of translation. It fits perfectly with the theme or discussion in an online media article. Because of the elaboration of this categorical syllogism is more easily understood by the readers (young child). This is in contrast to the hypothesis syllogism that requires antecedents to support its consequences and alternative syllogism that requires alternative propositions to support its conclusions. In daily life, the alternative syllogism is often used (searching address, when people does not find it in A place, it is likely to be in B place). However, in the writing of the online article, the syllogism is rarely used. CONCLUSIONS With argument reconstruction, there are many used types of arguments in online media articles. This research begins by determining the form of arguments in online media articles and analyzing argument reconstruction. This is evidenced by the number of data from both types of reasonings. From 5 articles, 44 paragraphs, and 150 sentences analyzed, 16 data are found to use inductive reasoning. About 11 data use generalization reasoning, and 4 data use cause-effect relationship. No data are found in the structural analogy, metaphorical analogy, and borrowed analogy. For deductive reasoning, 28 data are collected using categorical syllogistic reasoning. Meanwhile, no data are found related to the hypothesis and alternative syllogism. Based on the results of this research, it is found that the most commonly used argument form by online article writers is a generalization and categorical syllogistic reasoning. The cause-effect reasoning is also used although it is not many. REFERENCES Ardianto. (2015). Struktur argumen dalam wacana karya tulis ilmiah mahasiswa. Litera: Jurnal Penelitian Bahasa, Sastra, dan Pengajarannya, 14(1), 1−10. Bigi, S., & Morasso, S. G. (2012). Keywords, frames, and the reconstruction of material starting points in argumentation. Journal of Pragmatics, 44(10), 1135– 1149. https://doi.org/10.1016/j.pragma.2012.04.011 Cummings, A. L. (1999). Pragmatic. New York: Oxford University Press Inc. Eriyanti, R. W. (2017). Penalaran dalam tuturan lisan guru pada pembelajaran di SMP kota Malang. Litera: Jurnal Penelitian Bahasa, Sastra, dan Pengajarannya., 16(1), 78–95. Katzav, J., & Reed, C. (2008). Modelling argument recognition and reconstruction. Journal of Pragmatics, 40(1), 155–172. https://doi. org/10.1016/j.pragma.2007.07.004. Keraf, G. (2008). Argumentasi dan narasi. Jakarta: Gramedia Pustaka Utama. Mochales, R., & Ieven, A. (2009). Creating an argumentation corpus: Do theories apply to real arguments? A case study on the legal argumentation of the ECHR. In Proceedings of the 12th International Conference on Artificial Intelligence and Law (pp. 21−30). https:// doi.org/10.1145/1568234.1568238. 394 LINGUA CULTURA, Vol. 12 No. 4, November 2018, 389-394 Mundiri. (2016). Logika. Jakarta: Rajagrafindo Persada. Ridhani, A. (2013). Tipe argumen wacana argumentasi tulis siswa Sekolah Dasar kelas tinggi. Litera: Jurnal Penelitian Bahasa, Sastra, dan Pengajarannya, 12(1), 146−158. Suriasumantri, J. S. (2015). Filsafat ilmu sebuah apresiasi terhadap ilmu, agama, dan seni. Jakarta: Pustaka Sinar Harapan. Tusriawati, R., Saman, S., & Martono. (2014). Upaya meningkatkan kemampuan menulis paragraf argumentasi melalui metode STAD. Jurnal Pendidikan dan Pembelajaran, 3(12), 1−15. Van Den Hoven, P., & Yang, Y. (2013). The argumentative reconstruction of multimodal discourse, taking the ABC coverage of President Hu Jintao’s visit to the USA as an example. Argumentation, 27(4), 403−424. https://doi.org/10.1007/s10503-013-9293-z.