A French-English contrastive analysis of je dis ça, je dis rien and (I’m) just saying in Internet blogs and forums

Par Alexandre Dias Da Silva
Publication en ligne le 06 septembre 2019

Texte intégral

1Over the last few years, we seem to have seen a tremendous increase in the use of je dis ça, je dis rien and (I’m) just saying 1. Their omnipresence in oral speech especially2 – altougth perhaps more prevalent among young people – makes them worthwhile of further investigation. Interestingly, both expressions have come to be used ritually on social networks, either under their full forms or dedicated hashtags #JDCJDR and #IJS, making them a fertile ground for research. The phrases have, to date, received only little attention from linguists. However, they have been much written about informally by non-specialists tired of hearing them, and downgrading them to the status of “gap fillers”. Overall, they have been described as signs of the speaker’s cowardice and reluctance to take the full responsibility for his/her previous statement:

« Si j’étais toi, je me mettrais au régime. Enfin, je dis ça, je dis rien. » Ah ouais ?? Alors pourquoi on a l’impression que, malgré tout, tu l’as dit ? Abominable expression qui perdure chez les moins jeunes, cette périphrase qui semble mettre son utilisateur à l’abri de toute critique quant à la sortie assassine qu’il vient d’innocemment asséner devrait être passible d’une amende tant elle est absurde et hypocrite. (Adèle Bréau, « Je dis ça, je dis rien » et 200 autres expressions in-sup-por-tables !, 2013)

Saying, “I'm just saying,” puts a fire escape onto the end of a sentence. It lets you express a stern — even rude — opinion, but not really. You’re just saying. It invites the listener to discount what we’ve just heard, even as we’re reeling from it. (Scott Simon, “It’s Rude! It’s Crude! It’s Stupid! Just Sayin’”, NPR Simon Says, December 18 2010)

2However critical these observations are, they both clearly pinpoint the discourse-pragmatic stakes shared by the two phrases: they are part of a face-saving strategy, or “rhetorical backoff” – “the speaker reaffirms his or her commitment to the truth of what was just said but not to the implications that could be drawn from having made those claims” (Lee-Goldman, 2011: 77). However, a contextual analysis of data samples shows that the above observations need nuancing, as they only consider the prototypical case where the phrase follow a statement which is likely to cause offense to the hearer. Working on a corpus that I expressly compiled for this study, I found out that the two phrases are used not only to “discount” deliberately offensive or controversial utterances, but also in a more literal way, when the speaker is aware of the offensive or controversial character of his/her utterances but does not necessarily mean any harm, or does not necessarily want to impose his/her point of view.

3In this article, I first go through the key principles of the enunciative framework I will be working in, and present my corpus.

4I then pore over the link between form and meaning, looking at both morphology and syntax of the phrases, and contrast the conditions of their appearance within discourse.

5Lastly, I carry out semantic and pragmatic analyses on examples taken from my corpus: I show how the morphosyntactic specificities of the two phrases directly affect their pragmatic potentialities, even though they share an unvarying operation.

1. Theoretical background and corpus-based study

1.1. Culioli’s TOPE, a linguistic model for enunciative adjustment

6In Antoine Culioli’s Theory of the Predicative and Enunciative Operations (Théorie des Opérations Prédicatives et Énonciatives, in French), the enunciator (énonciateur) is the origin of the subjective location (repérages subjectifs). In this approach to language, markers (or traces énonciatives) – either implicit or explicit – of underlying, complex operations can be found in discourse, allowing to identify operations, which positing an origin (the enunciator) is part of.

7I will be working within the general framework of the Theory of the Predicative and Enunciative Operations (henceforth TOPE), as elaborated by Culioli. At its core lies the principle according to which linguistic items are markers of operations:

Language, which is a meaningful representational activity3, is only accessible through texts, that is, through patterns of markers which are themselves traces of underlying operations […] the goal is […] to re-construct, by a theoretical and formal process of a foundational sort, the primitive notions, elementary operations, rules and schemata which generate grammatical categories and patterns specific to each language. In short, the goal is to find invariants which found and regulate language activity, as this activity manifests itself through the meaningful configurations of different languages. (Culioli, 1990: 72)

8The speaker is the subjective origin committing himself/herself, i.e. endorsing the propositional content. French enunciativists traditionally see “commitment modality” (“modalité de la prise en charge”) as relating to the subjective origin validating or contemplating the validation of the propositional content (Celle, 2006; 56-57). To commit oneself to p is to declare that p is valid (or non-valid), or in other words, that p is the case (or not the case). Not to commit amounts to suspend the assertion. Commitment and assertion go hand in hand:

To be able to assert, to produce an assertion, I must declare it publicly […]; but there also needs be the commitment of someone who takes responsibility, who guarantees your statement or who wishes to assert something despite your position. If it is simply an instance of locution, you do not really have assertion. For there to be responsibility, one must guarantee what one states. In an institutional setting, the guarantor could be sanctioned if what he guarantees does not materialize. (Culioli 1995: 92 quoted by Celle, 2008: 17)

9Arguably, the content of an utterance cannot be limited to what is actually said: there is no certainty that what is actually said matches what is meant to be said. Assertion proves problematic owing to its very nature: asserting something can never be complete; there is always a gap between what the speaker means/wishes to say and what is actually said/understood. In order for communication to be somewhat productive, “speakers must find a way to evoke appropriate mental representations of the people, places, and things they want to talk about” (Jucker, Smith & Lüdge 2003: 1739) and mutual understanding is to be regarded as an operation of enunciative adjustment with both its successful and unsuccessful attempts. Any utterance may come to be challenged or contested, which is likely to lead the speaker to reformulate previous utterances in an attempt to clarify what s/he means. The speaker has several ways to do that, be it in “referring back” to – but also sometimes, announcing – a certain infelicitous word or in resorting to paraphrase.

As any operation that produces meaning, je dis ça, je dis rien and (I’m) just saying involve an operation: that of anaphora4, through anaphoric element ça and the anaphoric value conveyed by progressive aspect be + ing. In short, with these expressions, the speaker is performing a work revolving around re-(flexivity) and meta-(enunciation). (Vassiliadou, 2005)

1.2. Description of the corpus and adopted method of analysis

10To make up for the lack of existing French data, I based my work on a set of comparable data. My corpus is quantitatively and qualitatively homogeneous: it includes a French sample compiling 50 occurrences of je dis ça, je dis rien and an English one compiling 50 occurrences of (I’m) just saying. These instances were all retrieved from discussion groups, article comments, blogs, and forums: those are indeed written genres mimicking oral discourse, functioning on a conversational mode, and providing the needed background – confrontation of divergent opinions, propitious to informal, loosely organized debate – for the two colloquial expressions to occur. As they appeared only recently in both languages, I gathered occurrences on a limited time span spreading from 2010 to 2014 conversation threads.


Je dis ça, je dis rien

(I’m) just saying


44% (22)

50% (25)


26% (13)

32% (16)

Article comments

30% (15)

18% (9)

Table 1. Composition of the corpus

2. Je dis ça, je dis rien vs. (I’m) just saying: morphosyntactic considerations

2.1. Syntactic placement

11Were we to consider only the authors’ analyses that we reviewed in the introduction, then we should expect our markers to occur only after the problematic assertion (i.e. in final position) in order to modalize what has just been said: in the examples that the authors chose to illustrate their point, both markers occur at the very end. A quantitative analysis of my corpus shows that, indeed, both markers mostly occur as closing remarks, and that final position is prototypical. However, perhaps counterintuitively, I also found occurrences of both markers in initial position, as announcing markers (albeit in lesser proportions).


Je dis ça, je dis rien

(I’m) just saying


82% (41)

86% (43)


18% (9)

10% (5)

100% (50)

96% 5(47)

Table 2. Distribution of je dis ça, je dis rien and (I’m) just saying in my corpus (anticipatory uses only)

12Then, two configurations are possible for the markers to occur: either in final position, or in initial position – i.e. either before or after the element targeted by the marker. Accordingly, they may assume either a cataphoric function, or an anaphoric function.

2.2. Operations of anaphora/cataphora: ça vs. be + ing form

13Both expressions can be used cataphorically to achieve the same pragmatic effect than in final position:

(1) Moi je dis ça je dis rien, mais techniquement les petites annonces c'est réservé aux adhérents, et c'est pas la bonne section. Bref, mal placé.

(2) [Article comment] Just saying, but the phrasing of the name form on your application is terrible. I’d suggest not using “real name” and legal name interchangeably because it’s transphobic.

14But prototypically, both je dis ça, je dis rien and (I’m) just saying are to be found in final position, as punctuating or closing remarks. In that case, anaphoric reference is not achieved through the same medium for both expressions: whereas ça explicitly heralds the anaphora in je dis ça, je dis rien, it is the be + ing form that implicitly refers to what the speaker has just said in (I’m) just saying.

15Deixis is more often used in spoken language than it is in written discourse. In French, je dis ça, je dis rien illustrates this tendency with its demonstrative pronoun ça. Deixis also signals intersubjective involvement since the use of ça, in our case, is only possible because both the speaker and the hearer share common knowledge. Let us consider the following example and try to see what ça refers to:

(3) Bah c’est plutôt sur la longueur la différence. Avant de faire un roman, tu devrais commencer par une nouvelle. je dis ça, je dis rien

16Deictic pronoun ça is anaphoric since it points to the previous sentence (fléchage contextuel avec reprise du contexte à gauche, see e.g. Chuquet and Paillard, 1987 : 43). The speaker gives some advice (tu devrais) and then, the speaker withdraws. The expression can be fleshed out as follows: quand je dis ça, je dis rien > lorsque je dis ce que je viens de te dire, je ne dis rien > que je dise ça ou que je ne dise rien, cela revient au même. The last gloss evidences a problem at the level of assertion: what has been said cannot be “unsaid” by simply dismissing it. Clearly, the speaker’s resorting to je dis ça, je dis rien is an attempt to qualify and downtone his/her previous statement.

17In English, be + ing form implicitly refers to what the speaker has just said. Grammatical aspect be + ing allows for anaphora with commentary value on the immediate left context (Bouscaren, 1993: 20):

(4) [On a blog] Coke always tastes better out of the bottle, I don’t know why, Im just sayin’

2.3. Influence of morphology and syntax on meaning

18One major difference between the two markers is that, whereas (I’m) just saying maintains the assertion (though suspending it), je dis ça, je dis rien cancels it through explicit deassertion. In deasserting, the speaker marks his/her refusal to commit to his/her utterances along with the implications that could be inferred from having made those claims. This distinction is in part attributable to the semantic content of adverbs rien and just. Using je dis ça, je dis rien, the speaker altogether denies the validity or legitimacy of his/her assertion. In contrast, (I’m) just saying indicates that the speaker still acknowledges the validity or legitimacy of his/her assertion – but up to a certain limit, often to be determined by the interlocutor.

19In je dis ça, je dis rien, the discarding of the utterance is achieved by the semantic content of rien6: parataxis binds je dis ça and je dis rien together – the comma that is often to be found between the two segments somehow plays the role of an equals sign – and the identity of je dis (the common factor, so to speak) in both segments seems to imply that ça equals rien. Hence another gloss, which is applicable to any occurrence of je dis ça, je dis rien: j’ai dit ça, mais j’aurais très bien pu ne rien dire (ce serait alors revenu au même ; ce n’est jamais qu’un avis parmi tant d’autres).

20On the other hand, just in (I’m) just saying is a marker of a gap between what is meant to be said and what is actually said, and is central to the modalizing value of (I’m) just saying. It cannot be dropped and is not interchangeable with equivalent adverbs (?? I’m saying; ?? I’m only saying; ?? I’m simply saying). Culioli (1997) studied French marker seul(ement) to illustrate his concept of notion. Two properties surfaced from his analysis (our translation):

On the one hand, seul(ement) signals that we are dealing with a nonzero (“non nulle”) value (we are establishing the existence of x); on the other hand, seul(ement) signals that we have no more than x. Then seul(ement) conveys a double discontinuity: (nonzero/zero) and (more/no more). (Culioli, 1997: 16)

21Adapting Culioli’s French examples for seul(ement), we see that his remarks also apply to just:

(a) I’ve just 10 dollars.

(b) It’s just 5 pm.

22What is presupposed in (a) is that I have a certain amount of money which is nonzero, but I have no more than 10 dollars (a possible gloss is: perhaps it won’t be enough). What differentiates (a) from (b) though is that, while I can have no money (not even 1 cent), time must always be nonzero. In (b), just signals that the distance between actual time and estimated time is (i) nonzero (otherwise, the two would coincide and the utterance would no longer be relevant) and (ii) inferior to estimated time (let us say 6 pm) since it is not 6 pm yet, but it is not more than 5 pm either. Just 5 pm marks that it is not the estimated time and it is not more than the estimated/actual time. Therefore, the two invariants are nonzero-ness and no-more-ness. There is a non-coincidence, that is to say a variance (or discrepancy), going on between two elements that are not level. Just is the trace of an operation of the speaker’s acknowledgment of this unevenness.

23Now back to (I’m) just saying; with just, the nonzero value is taken into account in the sense that I could have said nothing but I have said this and just this: I am not saying less, I am not saying more either (theoretically). Just signals that (i) the distance is nonzero (otherwise, there would be coincidence between what is said and what is meant to be said) (ii) the speaker “worked out” that there was a non-coincidence between what s/he said and what s/he meant to say and his/her use of just marks that what is distinguished is inferior than intended (mes mots ont (peut-être) dépassé ma pensée).

24Lee (1999) identified that just can be used to hedge a statement, making it less peremptory (just a suggestion), but also have specificatory (just after 5 pm) and restrictive (just on Mondays) uses. This polysemy of just is central to account for the polysemy of the English marker and to determine its final meaning in context.

25Anaphora in je dis ça, je dis rien and (I’m) just saying is possible thanks to verbs dire and say, signaling that the speaker alters or adjusts his/her “way of saying”: s/he says his/her utterance in a “nonstandard” fashion and emphasizes the particular status of what s/he says as something being just said or something s/he equates to nothing. Verbs dire and say mean that the speaker is trying to bring what s/he has to say about the world within the reach of his/her interlocutor or, to put it differently, to make his/her internal representation about the world accessible to his/her interlocutor. In the following examples, the use of coordinating conjunction so is telling about how this alternative mode of saying that (I’m) just saying establishes “shields” the speaker from retaliation7:

(5) Just saying, so please, no offense intended.

(6) just saying so don't go getting all defensive on me

26Concerning syntax, I was unable to find any regular pattern that would correlate the position of the marker and its meaning. In fact, it turns out that the marker’s position at the macrostructural level has a greater influence on the interpretation of the marker than at the microstructural level – the position of the marker within an utterance having little influence on its meaning: both meaning and scope cannot be predicted only from its syntactic position and broader context is then needed to decide.

27Lastly, two planes of enunciation can often be distinguished in the context where the markers occur: a standard plane [1] and a meta-enunciative plane [2], which our markers are always part of. Interestingly, co-occurring discourse markers8 can be used to signal and articulate the transition from one plane to another:

(7) [Tu peux effectivement réfléchir à une transition. Sinon tu peux aussi1 devenir plus "garçonne" et peut-être que ça comblera en partie tes "manques".]1 Enfin [j’dis ça, j’dis rien]2

(8) [It’s always best to start with your GP for a referral. Also many non med things can be done to lose weight, as many are unaware of the body's amazing ability to burn fat (not calories) all on it’s own, my manipulating what we intake, when we intake and how much. Eliminating all forms of sugar, artificial sweeteners, caffeine, gluten, proper hydration, etc. all help the metabolism burn fat. Most meds may help a little, more than likely the weight comes back because it's the lifestyle change that keeps it off.]1 Anyway, [just sayin']2

3. Corpus study and semantico-pragmatic considerations

28Modality markers allow us to make utterances acceptable and efficient at the same time. The speaker reasserts the continuity of his/her discourse, indicating that s/he (more or less) maintains his/her standpoint: they are “fine-tuning devices that produce a compromise between what the speaker wants to say and what the hearer is willing to accept” (Schneider, 2007: 97). Adapting one’s discourse (or “tuning in”) to the interlocutor in reconsidering otherness is essential to efficiently regulate and lubricate social relationships.

3.1. Pragmatic analysis: subjectivity markers. Defining invariants

29When using one of the markers, subjectivity surfaces on three planes. Firstly, the mere fact of speaking is an indicator that the speaker cares about what s/he is saying: the interlocutor’s not knowing how much seriousness (or detachment) s/he should attach to the speaker’s discourse is what the speaker plays on, and irony precisely exploits this gap. Secondly, an intersubjective space is set up: for instance, a relationship of dominance or authority between the speaker and his/her interlocutor may be established. Urmson (1963) argued that parentheticals could be used to alter or weaken one’s own claim to truth implied by an assertion: lack of hedging might cause speakers to come across as arrogant because of an overweening claim to truth, which is likely to lead to conflict. Thirdly, the use of modality reflects how the speaker positions himself/herself with regard to the predicative relation.

30I will now define the invariant for each marker, starting with what both have in common.

31Metalinguistic markers je dis ça, je dis rien and (I’m) just saying allow the speaker, who spontaneously presents himself/herself behind an altruistic/well-intentioned façade, to modalize his/her (subjective) said (“son dit”), thus diminishing his/her commitment to the propositional content of the utterance, yet signing9 it, i.e. acknowledging it as his/hers10, in spite of a pre-constructed concession relating to the partiality of the saying (“le dire”). This concession can be expressed as follows: I may say/have said something, but it will always be and remain something that I think(thought)/believe(d)/know(knew) is/was the case from my (then-)current standpoint and experience – within a very specific situation of utterance –, its interpretation also being dependent on my way of wording it (ce n’est jamais que ma façon de voir les choses à cet instant t, depuis ma configuration/disposition mentale actuelle, à la lumière de mon propre passé, de ma propre expérience et de ma façon de l’exprimer qui m’est propre ; du fait qu’il existe autant de variables, je me garde la possibilité de me raviser et de revenir sur mes propos ; aussi, je vous demande de vous y fier comme bon vous semble). As shown by this last sentence, the markers contain a pragmatic indication on how the speaker positions himself/herself with respect to his/her saying and an interlocutor-oriented pragmatic instruction11 on how his/her text should be/is to be read – the speaker promotes a flexible, deformable reading, not necessarily targetting a particular referent word/sequence of words but often addressing and altering the overall interpretation of the said: the interlocutor’s position as the ultimate judge of whether the speaker’s utterance is the case/relevant is reasserted, the speaker’s assertiveness having endangered it – the speaker evades any divergent standpoint and deflect counterarguments12; reasserting his/her interlocutor’s freedom to (dis)regard his/her assertion, the speaker is “shielded” from retaliation. Being aware of the burden to shoulder that assertion represents, the speaker – assessing the (in)significance of his/her saying resulting from the selection of p as compared out of p’ in a meta-enunciative movement – allows himself/herself some latitude (that can be more or less broad, and sometimes minimal) for possible future adjustments: otherness is (more or less) maintained.

32Their morphosyntactic differences, however, cause them to have slightly different pragmatic effects. I will now continue to define both invariants independently:

Je dis ça, je dis rien

(I’m) just saying

Deassertion (rien)

Semantic content of rien induces the speaker’s denying the validity of one’s own assertion altogether and suspending the belief in something13: “something is the case” but something else is not quite the case, and the speaker ceases to endorse what is said

  • Refusal to commit

Partial deassertion (just)

Semantic content of just induces the creation of a boundary14/limit of the acceptable: understand what you should understand and only/just what you should understand, not more (up to the point of what you consider to be correct/appropriate/desirable)

  • Minimal commitment

Table 3. Influence of morphology and syntax on pragmatics

3.2. Classifying occurrences in terms of modality: polysemous markers

33All occurrences of je dis ça, je dis rien and (I’m) just saying from my corpus are classifiable according to Culioli’s typology of modalities – in this section, I take some occurrences from my corpus that I classify and analyze in terms of modality and meaning:

34- Type I modality (modality of assertion) concerns the speaker’s “[defining] propositional content, as it is represented by the predicative relation, as being validated, that is true or false” (Bouscaren, Chuquet & Danon-Boileau 1992 :40). This includes not only positive or negative assertions, but also interrogations and injunction. This type of modality does not apply to our occurrences.

35- Type II modality (epistemic modality) “establishes a relation between the utterer and a propositional content as represented by the predicative relation” (Bouscaren et al. 1992: 37). Here, the probability of validation of the predicative relation is (quantitatively) assessed without making a choice between validation and non-validation. Enunciative commitment is minimal and the speaker cannot be decisive on the validity of the predicative relation.

(9) Je pense qu'une personne plus expérimentée que moi pourra répondre à ce sujet. Je suggère que ça vient d'un problème de port mais après j'dis ça, j'dis rien.

(10) [On a tech forum] Je pense plutôt à un bug dans le catalogue du disque du serveur : il est très étonnant que le serveur ait accepté que deux fichiers (l'indestructible et le plus récent) portant le même nom. De plus, j'ai essayé de rapatrier le fichier indestructible sur mon disque, pour voir ce qu'il contenait. Impossible, le fichier n'existe pas. Je pense qu'il avait été détruit, mais que le catalogue du disque du serveur n'avait pas été mis à jour. Enfin, je dis ça, mais je ne dis rien: ce n'est pas moi l'expert.  ;D

(11) Just tell them (politely, please) to have them take your number of their list and they'll do it. If you get another call from NRG then that person probably isn't doing their job correctly or accidentally put it as something else (it happens.) Again, I don't know if it actually works, just sayin’.

(12) [About cameras] A new focus system which includes 9 cross-type focus points vs just one for the T3i is another impressive improvement. It would seem that this one item will push the T4i to a level playing field with the 60D.. just sayin’. (I’m no expert, just a casual observation)

36Observing the context in which just sayin’ occurs in (12) allows us to spot several hints of deassertion and deactualization: it would seem and modal auxiliary will. Just sayin’ comes right after to support this movement of deassertion. The speaker also adds an extra sentence after just sayin’ so as to justify this deassertion: I’m no expert, just a casual observation. Similarly in (10), we have ce n’est pas moi l’expert. Just sayin’ could be glossed in: I’m just making a casual observation. We are dealing with epistemic modality since the speaker does not really know whether the predicative relation is verified or not.

37- Type III modality (appreciative modality) concerns a qualitative judgment expressed by the speaker regarding the content of the predicative relation (i.e. it is <good, bad, normal, strange> that…).

(13) C'est bidon comme argument : il est toujours possible de mentir, tu te fais passé pour un hétéro et tout va bien. C'est même plus bidon que ça, ça encourage davantage le mensonge, en plus de préjugés légèrement homophobes... m'enfin j'dis ça, j'dis rien.

(14) Avant qu’elle soit au gouvernement, Taubira faisait partie d’un groupe anti-blancs. C’est clair, elle est raciste. J’dis ça, j’dis rien, mais c’est pitoyable de retourner sa veste comme ça

(15) Just saying, but the phrasing of the name form on your application is terrible. I'd suggest not using "real name" and legal name interchangeably because it's transphobic.

(16) To be fair, unless you've played the game it's difficult to judge the combat system. (Not knocking your right to contribute feedback, just saying so don't go getting all defensive on me Image 10000201000000110000001297AF56B85E091DCF.png)

38Here, the phrases introduce and modalize an evaluative statement, which is related to what the speaker’s normative judgment: bidon, pitoyable, terrible… Modalization conveys the speaker’s reluctance to fully commit and take the full responsibility for his/her statement. In (14), the marker can be equated to: je m’apprête à dire quelque chose, ce quelque chose pourra vous sembler peu pertinent voire déplaisant, mais je le dis quand même car c’est mon avis. Hence the mais15 that appears after the expression. This is observable in (15) too with but: just saying, but here is my take on the subject. In (14), this specific use of j’dis ça, j’dis rien is reminiscent of another expression, je veux pas dire, and both are completely interchangeable here: it has to do with something the speaker is not ready to accept the full responsibility for.

39- Type IV modality (root modality, or intersubjective modality) “concerns the relation between the grammatical subject of the utterance and the predicate” (Bouscaren et al. 1992: 37). The speaker is trying to have an influence on others, pressuring his/her interlocutors into doing something – hence the intersubjective aspect of this modality, the relation between speaker and hearer being privileged (although intersubjective value is not exclusive to type IV modality).

(17) Faire un tableau « les cadeaux de mes rêves » quelques semaines avant son anniversaire peut s’avérer très utile par exemple, moi j’dis ça, j’dis rien...

(18) J'dis ça, j'dis rien, mais j'le dis quand même, mais on a déjà des chevaux et des licornes...donc pourquoi pas des lamas ?

(19) [On a blog] Spoiler Alert: If you haven’t seen Fate/Zero and you’re looking to protect yourself from plot spoilers, you probably shouldn’t keep reading. Just sayin’!

(20) You didn't get the point, you don't charge a cars battery while its hooked up because power is still being drawn. Same with the phone, best to turn it off and charge. Just saying though, you don't have to feel like you need to do it. I do it at night while I sleep and get great battery life.

40The meaning of the markers here is I don’t want to compel you to do it, but personally, this is what I (would) do (and this is what works for me, so it should work for you). This is exactly what we have in example (20): I do it at night while I sleep and get great battery life. but you don’t have to feel like you need to do it. Again, it is interesting to note that in (17), we have verbal modalization with peut along with verb faire in its infinitive form, so modalization works in synergy with deactualization. Therefore, the phrases participate in the modalization process initiated in their respective sentences since they allow the speaker to deassert and not to sound like a moralizing force.

3.3. Anticipation and adjustment

41What my corpus-based study further revealed is that another configuration is possible for (I’m) just saying, corresponding to the last line of the table below: another major difference between the French and the English marker is that (I’m) just saying can also be used in palliative adjustment, in addition to its anticipatory use. In that case, the speaker tries to save face when s/he considers that his/her interlocutor’s reaction is unfitting or unexpected. Using (I’m) just saying would supposedly bring immediate relief to defuse the situation, allowing to counterbalance the rudeness, gaucherie, etc. of the previous utterance.


Je dis ça, je dis rien

(I’m) just saying



82% (41)

86% (43)


18% (9)

10% (5)

100% (50)

96% (47)


4% (2)

Table 4. Distribution of je dis ça, je dis rien and (I’m) just saying in my corpus

3.3.1. (I’m) just saying: anticipation + adjustment

42To sum up, (I’m) just saying may be used either:

43- in anticipation, as a pre-emptive measure, when there is a risk of being misunderstood and negated: the speaker attempts to forestall conflict/disagreement,

44- or in palliative adjustment when conflict/disagreement is overtly declared, as a neutralizing phrase: the actual effect of the speaker’s previous utterance on his/her interlocutor proved different than expected; the interlocutor “failed” to gauge/weigh the speaker’s words.



Initial position

(21) Just saying, but the phrasing of the name form on your application is terrible. I’d suggest not using “real name” and legal name interchangeably because it’s transphobic.

(23) A: I think u should get better at the variel flip before you teach others cus u look like ur not very comfortable on the board when u do that trick and u go to slow

B: hey faggot stfu16 cunt.

A: hey chill man im just sayin dont get mad

Final position

(22) Coke always tastes better out of the bottle, I don’t know why, Im just sayin’

Table 5. (I’m) just saying: anticipation vs. adjustment

45Furthermore, in the case of adjustment, (I’m) just saying is necessarily anaphoric. In (23), the marker has scope over A’s first turn. Deassertion is achieved in the aftermath of B’s unexpected and unnecessarily coarse reaction. As can be seen from the recapitulative table above, there are hardly any occurrences of (I’m) just saying used in adjustment in our corpus (4%). One reason for this might be that discussion is not instantaneous on online platforms and that answers are delayed. As a result, it would not make sense to use (I’m) just saying in adjustment a few hours afterwards: to effectively “ease the burn,” (I’m) just saying has to occur immediately after the addressee’s unfitting reaction.

3.3.2. Je dis ça, je dis rien: anticipation only

46In my corpus, je dis ça, je dis rien only occurs in cases of anticipation. One of the reason for this is the incompatibility of French simple present and adjustment, which is made obvious when we gloss the marker: lorsque je dis ce que je viens de te dire, je ne dis rien and not lorsque j’ai dit ce que je t’ai dit, je ne disais rien. Moreover, searching for occurrences of je disais ça, je disais rien (or je disais ça, je dis rien) on Google does not prove very fruitful: the few hits that we get are not very convincing and remain marginal. French imparfait or passé composé appear to be more appropriate in this context, yet je disais ça, je disais rien or j’ai dit ça, j’ai rien dit still sound strange. Rather, equivalent French counterparts in cases of adjustments could be je disais ça comme ça or j’ai rien dit.


47Je dis ça, je dis rien and (I’m) just saying definitely fall into what Jamet & Jobert (2001 : 11) call “déodorant[s] du langage.” A compositional analysis of the two markers showed that their constituents had a direct impact on their pragmatic potentialities; yet their pragmatic effects remain highly similar. The main difference in discourse is that our French marker can only be used in anticipation, and not in retroactive adjustment. Although there are many distinct uses of the two markers, they both fall within the same problem of apprehending the interlocutor’s reception of the speaker’s assertive utterance.

48As most pragmatic forms, our phrases do not contribute to propositional meaning and their scopes are not intrasentential: they play a role on the macro-textual level and fulfill semantico-pragmatic functions. They also appear to have reached a high degree of fixation: subjective markers je and I cannot be replaced with another grammatical person (*tu dis ça, tu dis rien/*you’re just saying). Dostie (2004) notes that first-person fixation is the trace of the interactants’ presence involved in the enunciative process. This is not surprising, since discourse markers are considered as traces of the speaker’s subjectivity in texts: they are to be seen as evidence of speaker commitment with respect to discourse content insofar as they are part and parcel of the building of the enunciative stage, as suggested by Paillard (2009). (I’m) just saying has reached a high level of pragmaticalization, though not maximal, and je dis ça, je dis rien appears to be on its way towards ridding itself of its heavy paratactic structure: its shorter version je dis ça… seems to be emerging recently.

49Having analyzed the different meanings of the markers and their pragmatic impacts allows me to conclude that both markers, in spite of their morphosyntactic dissimilarities causing (I’m) just saying to be more pragmaticalized – thus more flexible – and to be eligible for both anticipation and adjustment, are satisfactory equivalents in cases of anticipation.


Andersen, H. L., 2007, « Marqueurs discursifs propositionnels », Langue française n° 154(2), p. 13-28.

Bouscaren, J. & Chuquet, J., 1987, Grammaire et textes anglais, Guide pour l’analyse linguistique, Paris – Gap, Ophrys.

Bouscaren, J., 1993, Linguistique anglaise : initiation à une grammaire de l’énonciation, Paris – Gap, Ophrys.

Bouscaren, J., Chuquet, J. & Danon-Boileau, L., 1992, Introduction to a Linguistic Grammar of English. An Utterer-Centered Approach (R. Flintham & J. Bouscaren, Trans.), Gap, Ophrys.

Brinton, Laurel J., 2005, “Processes Underlying the Development of Pragmatic Markers: The Case of (I) Say”, in Opening Windows on Texts and Discourses of the Past, Amsterdam, John Benjamins Publishing, p. 279-99.

Celle, A., 2006, Temps et modalité. L’anglais, le français et l’allemand en contraste, Bern, Peter Lang.

Celle, A., 2008, “Tense, Modality and Commitment in Modes of Mixed Enunciation” in P. De Brabanter & P. Dendale (dir.), Commitment. Belgian Journal of Linguistics n° 22, Amsterdam, John Benjamins, p. 15-36.

Culioli, A., 1990, Pour une linguistique de l’énonciation : opérations et représentations (Tome 1), Paris – Gap, Ophrys.

Culioli, A., 1995, Cognition and Representation in Linguistic Theory, Amsterdam, John Benjamins.

Culioli, A., 1997, « À propos de la notion » in C. Rivière & M.-L. Groussier (dir.), La notion : actes du colloque “La notion” tenu à l’UFR d’Études anglophones, Université Paris 7 - Denis Diderot, en février 1996, Paris – Gap, Ophrys, p. 9-24.

Culioli, A., 1999, Pour une linguistique de l’énonciation : domaine notionnel (Tome 3), Paris – Gap, Ophrys.

Culioli, A., 2001, « Heureusement! » in Saberes no tempo–Homenagem a Maria Henriquez Costa Campos, Lisbonne, Edições Colibri, p. 279-284.

Dostie, G., 2004, Pragmaticalisation et marqueurs discursifs : analyse sémantique et traitement lexicographique, Brussels, De Boeck.Duculot.

Groussier, M.-L. & Rivière, C., 1996, Les mots de la linguistique : lexique de linguistique énonciative, Paris - Gap, Ophrys.

Jamet, D. & Jobert, M., 2010, « Juste un petit mot sur l’euphémisme... » in D. Jamet & M. Jobert (dir.), Empreintes de l’euphémisme : tours et détours. Paris, L’Harmattan, p. 11-28.

Jucker, A. H., Smith, S. W. & Lüdge, T, 2003, “Interactive aspects of vagueness in conversation”, Journal of Pragmatics, n° 35(12), p. 1737-1769.

Hansen, Maj-Britt Mosegaard, 1998, « The Semantic Status of Discourse Markers », Lingua 104 (3–4), p. 235-60.

Hopper, Paul J., 2003, Grammaticalization, 2nd ed., Cambridge/New York, Cambridge University Press.

Hübler, A., 1983, Understatements and Hedges in English, Amsterdam, John Benjamins.

Lee, D. A., 1991, « Categories in the description of just », Lingua, n° 83(1), p. 43-66.

Lee-Goldman, R. R., 2011, Context in Constructions, PhD Dissertation, University of California, Berkeley.

Paillard, D., 2009, « Prise en charge, commitment ou scène énonciative », Langue française, n° 162(2), p. 109-128.

Rossari, C., 1993, « A propos de l’influence de la composition morphologique d’une locution sur son fonctionnement sémantico-pragmatique », Cahiers de linguistique française, n°14, p. 151-171.

Schiffrin, D., 1987, Discourse Markers, Cambridge, Cambridge University Press.

Schneider, S., 2007, Reduced Parenthetical Clauses as Mitigators: A corpus study of spoken French, Italian and Spanish, Amsterdam, John Benjamins.

Secova, Maria, 2011, Discourse-Pragmatic Features of Spoken French: Analysis and Pedagogical Implications, PhD Dissertation, Queen Mary, University of London.

Urmson, J. O., 1963, « Parenthetical Verbs » in C. E. Caton (dir.), Philosophy and Ordinary Language, Urbana, University of Illinois Press, p. 220-246.

Vassiliadou, H., 2005, « La reformulation : Marqueurs linguistiques, stratégies énonciatives » in M.-C. L. Bot, M. Schuwer, & E. Richard (dir.), Pragmatique de la reformulation, types de discours, interactions didactiques, Rennes, Presses Universitaires de Rennes, p. 35-50.


1 Use of I’m just saying in the spoken component of the Corpus of Contemporary American English increased twofold from 1990 to 2012.

2 After establishing descriptive and inferential statistics from the Corpus of Contemporary American English (464,020,256 words at the time of my research), I found out that I’m just saying is approximately 7 times more frequent in speech than in writing.

3 Emphasis mine

4 I use anaphora as a substitute for the notion of reprise.

5 See Table 4 below for the remaining 4%.occurrences.

6 Rien can not only mean “néant” but also “peu de chose” (cf. substantival use: “un rien”) > je dis rien je dis pas grand chose ce que je dis n’a pas ou peu de valeur

7 I noticed the same thing in French: Je dis ça, je dis rien alors tu devrais attendre la réponse de quelqu'un de plus expérimenté que moi !

8 Enfin, mais, bon, bref, après (tout), voilà and anyway, but, (oh) well, whatever, though are the most frequent and are likely to combine. Incidentally, après (tout), anyway, and though seem to be markers specialized in reintroducing otherness.

9 “les MDP [marqueurs discursifs propositionnels] à la première personne sont plus fréquents dans les situations de discussion libre à plusieurs locuteurs, où on pourrait dire qu’il y a un besoin de « signer » l’énoncé ; dans les situations d’interview, les locuteurs se voient attribuer par définition la parole et ont moins besoin d’expliciter la subjectivité de leurs énoncés.” (Andersen, 2007, p. 19)

10 The assertive source is identified: je/I.

11 “Preventing the hearer from negating the propositional content is meta-communicative in that the primary communicative function of asserting something is accompanied by instructions as how to process the assertion.” (Hübler, 1983, p. 148)

12 “[…] toute assertion écarte autrui en tant que contestataire éventuel : ce qui est asserté par un engagement subjectif est stabilisé par la force du statut social, du recours à une norme éthique, à une procédure rationnelle, vis-à-vis de toute éventuelle polémique.” (Culioli, 1999, p. 159).

13 “Une assertion est le produit d’une double opération : (1) d’un côté on dit que « quelque chose est le cas », c’est-à-dire que un « quelque chose » est repéré par rapport à un système de référence […]. Un cas se définit comme un état de choses repéré […]. Dire que « quelque chose est le cas », c’est situer ce quelque chose (une occurrence d’une relation prédicative non encore identifiée) dans un espace référentiel ; (2) d’un autre côté, on construit une occurrence (en suspens) d’une relation prédicative spécifique, en attente d’un éventuel repérage de validation (ce que j'appelle une lexis) […] Cette opération de validation (sélection de la valeur estimée adéquate par un sujet) s’accompagne de deux opérations d’ordre subjectif : (1) engagement (≠ commitment dit D. Paillard) du sujet énonciateur qui, par delà son activité de locuteur, tient à dire (rendre accessible à autrui) ce qu’il sait, pense, croit être le descripteur adéquat ; (2) valuation par le sujet énonciateur qui assigne une valeur téléonomique (avantageux, désavantageux…) à l’asserté.” (Culioli, 2001, p. 280).

14 Just creates a boundary which, if crossed by the interlocutor, automatically brings about commitment reversal: the speaker is no more responsible for his/her utterances and it is the interlocutor who now has to shoulder the responsibility for his/her overinterpreting the speaker’s utterances.

15 Je dis ça, je dis rien relies on a performative contradiction. Concessive conjunctions mais and but signal a disruption in the expected course of events, allowing modulation of stancetaking: “Le terme concession vient de ce que l’énonciateur qui prend ainsi en charge ce qui peut être considéré comme une contradiction, le fait en connaissance de cause, c’est-à-dire en concédant au co-énonciateur l’existence de cette contradiction.” (Groussier & Rivière, 1996). Our markers tend do minimize the importance of the speaker’s utterances, even when the speaker deems to have made a point (the interlocutor may then be hesitating between a euphemistic or a litotic reading of the speaker’s utterances). Resorting to je dis ça, je dis rien or (I’m) just saying, the speaker seems to indicate that s/he is just saying something in passing and that s/he does not want to delve further in. In some contexts, je dis rien could in fact be understood as je le dis haut et fort.

16 Shut the fuck up

Pour citer ce document

Par Alexandre Dias Da Silva, «A French-English contrastive analysis of je dis ça, je dis rien and (I’m) just saying in Internet blogs and forums», Cahiers FoReLLIS - Formes et Représentations en Linguistique, Littérature et dans les arts de l'Image et de la Scène [En ligne], Cahiers en ligne (depuis 2013), Traces de subjectivité et corpus multilingues, III. Les données des corpus comparables : comparaison des langues dans des genres ciblés, mis à jour le : 23/04/2020, URL : https://cahiersforell.edel.univ-poitiers.fr:443/cahiersforell/index.php?id=697.

Quelques mots à propos de :  Alexandre Dias Da Silva

University of Paris Diderot – Sorbonne Paris Cité, CLILLAC-ARP EA 3967