Barbara Prem
Section 438(2) of the Bharatiya Nagarik Suraksha Sanhita, 2023 (BNSS), formerly Section 397(2) of the CrPC, bars the revision of interlocutory orders to prevent trial delays. However, inconsistent judicial application of the ‘intermediate order’ doctrine, established in Madhu Limaye v. State of Maharashtra, has created a classification crisis. By miscategorising crucial mid-trial orders such as the refusal to summon defence witnesses as purely interlocutory, courts routinely shield them from revision. This formalistic classification directly undermines the accused’s substantive right to a fair trial under Article 21, prioritising procedural expediency over procedural fairness.
This piece argues that the core procedural uncertainties and a lack of fair trial guarantee for the accused under S. 438(2) BNSS stem not from the mere lack of proper definition or from the existence of a revision bar on interlocutory orders. Rather, they stem from misinterpretations of the ‘intermediate order’ concept laid down in the landmark judgement Madhu Limaye and the restricted interpretation of ‘interlocutory order’.
Firstly, the piece traces the evolution of the provision. Secondly, it discusses the attempts made by the judiciary to define interlocutory orders and tests laid down, including the landmark Madhu Limaye case and the concept of ‘intermediate order.’ Thirdly, it discusses two categories of misinterpretations of the Madhu Limaye case by the judiciary and their implications. Fourthly, it discusses the interplay of S. 438(2) BNSS with S. 528 BNSS. Lastly, it shows the correct interpretations of the case and its subsequent guarantee of a fair trial to the accused. The cases used for discussing misinterpretations and right interpretations pertain to summoning of witnesses, accused, and producing documents.
Background of the provision
The revision power in S.435 of the CrPC 1898 made no distinction for the exercise of revisionary powers with respect to interlocutory and other orders, thereby allowing revision even for purely interlocutory orders. The 41st Law Commission Report (1969) suggested changes in powers of High Courts, including revision power, to prevent delays, while upholding basic considerations of ensuring fair trial and fair deal for poor sections of society. Paragraph 5(d) of the Statement of Objects and Reasons read, “the powers of revision against interlocutory orders are being taken away, as it has been found to be one of the main contributing factors in the delay of disposal of criminal cases”. Therefore, S. 397(2) of CrPC was introduced to prevent delays and ensure a fair trial. In reality, the accused’s right to fair trial is violated under this provision.
The BNSS introduced specific terminological shifts such as the amendment of Section 438(1) BNSS to exclude the general term “bail”, thereby to align with the new, distinct definitions in Section 2(1) BNSS. However, the actual statutory bar on revision in Section 438(2) remains identical to its predecessor. Since the BNSS still does not statutorily define “interlocutory order,” the jurisprudential burden of classification remains entirely on the courts.
No Definition, Many Tests: The Interlocutory Order Puzzle
While there is no explicit definition of interlocutory orders provided in CrPC or BNSS, courts have attempted to define and provide features of interlocutory orders. S. Kuppuswami Rao v. The King laid down the Finality test wherein an order can be said to be a final order only if, no matter whichever way the decision is given, the action will be determined. This test is problematic when applied to cognisance orders. This is because even if a court takes cognisance illegally or without jurisdiction, the order would be deemed interlocutory rather than final, thereby barring revision.
The Court in Mohan Lal Magan Lal Thacker v. State of Gujarat discussed 4 questions generally used by English Courts to decide the nature of an order. The questions are as follows;
(i) Whether the order made upon an application is such that a decision in favour of either party would determine the main dispute?; (ii) Whether the order was made upon an application upon which the main dispute could have been decided? (iii) Whether the order as made determined the dispute? (iv) If the order in question is reversed, whether the action would go on?
This test is insufficient since it focuses only on the impact of the order and not the risk of substantive right violations of the accused.
In 1977, the Court in Amarnath v. State of Haryana (“Amarnath”) provided that interlocutory order under S.397(2) CrPC has to be used in a restricted sense as it merely denotes orders of a purely temporary nature and does not affect important rights and liabilities of the parties. While it observes that an order which substantially affects the rights of the accused or decides certain rights of parties are not interlocutory orders, it maintains that summoning witnesses is an interlocutory order, even though witnesses can overturn the case and thereby affect the accused’s rights. This is because according to Amarnath, orders that are steps taken in aid of the pending proceeding, or steps taken towards the final adjudication and for the parties to assist in the prosecution of their case in a pending proceeding, and which only regulate the procedure are considered not to affect any rights or liabilities of parties. The judgement, therefore, fails in this point.
Madhu Limaye and V.C. Shukla: Shift towards Proper Tests
In the same year, the Supreme Court in Madhu Limaye was tasked with deciding whether a lower court’s order rejecting a preliminary objection was strictly interlocutory. The Court focused on the core procedural dilemma, i.e., whether an order that does not finally conclude a trial can still be subject to revision.
A key takeaway from this case is that the binary of interlocutory and final order does not hold true in all cases. The case uses interpretations of CPC to distinguish ‘intermediate orders’ and ‘case decided’ to show how an order may not be purely interlocutory. The difference is explained through an example of a defendant raising a plea of jurisdiction of a particular court to try a suit or bar limitation where if he succeeds, the action is determined finally in the court and if it fails, the suit proceeds. A point raised may be interwoven with other issues and it may not be decided as a preliminary point of law. Therefore, it may neither be interlocutory nor final.
Madhu Limaye also critiqued Kuppuswami’s finality test and Mohan Lal’s reliance on four questions, especially the fourth question, stating how both tests led to opposite conclusions. Applying the finality test, order of the Sessions Judge in the present case would not be a final order since if the order is declared invalid, the proceeding gets determined but if the order is declared valid, the defamatory proceedings will continue. According to the fourth question in Mohan Lal, “If the order in question is reversed, whether the action would go on?,” it would be a final order since the reversing of the Session Judge’s order would terminate the proceedings.
To solve such inconsistencies, the case introduced the concept of ‘intermediate orders’ and defined it as an order which prima facie seems to be interlocutory, but when reversed has the effect of terminating the proceedings thereby resulting in a final order. The court further clarified that an interlocutory order does not invariably mean the converse of the term final order, and that intermediate orders can be subject to revision.
Two years later, a 4-judge bench in V.C. Shukla v. State through CBI reiterated that intermediate, quasi-final and final orders were revisable. Contrary to Amarnath, it held that the term interlocutory order has to be given a “very liberal interpretation in favour of the accused” to ensure a completely fair trial.
This piece keeps the concept of intermediate order and liberal construction of the term ‘interlocutory order’ as yardsticks for comparison since they have been laid down by landmark cases and will ensure that the object of the provision, i.e., prevent delays while ensuring fair trial, is upheld.
Misinterpretations of Precedents
There are two types of misinterpretations of precedents: false dichotomy fallacy and selective reading. False dichotomy fallacy occurs when rejection of one proposition is taken to mean the truth of another proposition, without room for alternatives. By selective reading, courts have blatantly failed to acknowledge the rights of the accused being affected, focusing only on the definition of interlocutory orders as being ‘temporary’ in nature.
1. False Dichotomy Fallacy
In Girish Kumar Suneja v. CBI, the court held that revisional powers can be exercised only against a final order or an intermediate order. It rightly interprets an intermediate order as one which if passed in a certain way, the proceedings would terminate but if passed in another way, the proceedings would continue. The problem with such interpretation is that it has wrongly understood the test laid down in Madhu Limaye case to be exhaustive for determining what is an interlocutory order, rather than what is not. Girish Kumar Suneja used Madhu Limaye’s negative test as the only possible test and has strictly classified types of orders as interlocutory, intermediate and final. When an order is said not to be intermediate or final, it is taken to mean that the order is interlocutory and revision is barred. It ignores the requirement of following due process and ensuring that the accused’s rights are not violated, thereby prioritising formal categorisation over justice.
Such interpretation has drastically narrowed down the scope of intermediate orders to an order that would result in the culmination of proceedings if set aside. This would lead to neglect of mid-proceeding orders that substantially impact rights but do not terminate proceedings due to their classification as interlocutory orders. Examples include orders summoning witnesses, producing documents, etc. Such denial of defence evidence also violates Article 21 of the Constitution providing fair trial guarantees.
2. Selective Reading
In Sethuraman v. Rajamanickam, the trial court refused to call for documents and summon witnesses, which was reversed by the High Court and upon appeal, the Supreme Court held that the High Court could not have revised the trial court’s order since production of documents and summoning of witnesses were interlocutory orders.
Witness testimonies and documents as evidence often heavily influence the outcome of a trial, even if they do not immediately terminate the proceedings. This judgement narrowly interpreted interlocutory order as something that is not final, omitting to read the substantive rights of the accused into the definition, thereby disregarding such rights. It undermined the very safeguard Madhu Limaye provided.
Ten years later, M Koteswara Reddy v. State of Andhra Pradesh had an opportunity to refer Sethuraman back to the CJI and have it revisited by a larger bench. Rather, it decided that the interlocutory nature of orders summoning witnesses and documents was well settled, ignoring both Madhu Limaye and V.C Shukla.
Such misinterpretations by courts result in neglecting the rights of the accused for a fair trial and a fair deal for the poor sections of society. This fails the basic considerations laid down by the 41st Law Commission report in suggesting changes to prevent delays in order to ensure a fair trial.
Position of §528 BNSS with respect to Interlocutory and Intermediate Orders
S. 528 of the BNSS grants inherent powers to High Courts to quash criminal proceedings in the interests of justice and prevent abuse of process. Here, the question arises as to whether courts can revise interlocutory orders under S. 528.
Amarnath held that S. 482 CrPC contains only inherent powers of the court and does not confer ‘new powers’ and concluded that S. 482 will not apply to an order which is expressly barred from revision under S. 397(2) of CrPC. Further, it provides that orders which are matters of moment and which affect the rights of the accused shall be within the purview of revision. Madhu Limaye slightly modulates Amarnath to hold that revision under S. 482 shall be allowed if there is an abuse in process of law or in the interests of justice.
Therefore, intermediate orders can be subject to revision under Section 438 BNSS, failing which, Section 528 BNSS can be invoked as a safety valve.
Corrective Precedents Restoring the Rights-based Reading
While misinterpretations by courts created uncertainties, recent cases have rightly interpreted the intermediate order concept and restored the rights of the accused, thereby ensuring that the object of S. 397(2) is upheld.
In 2017, the Allahabad High Court in Sanjay Singh v. State of UP held that summoning an the accused was an intermediate or quasi-final order and not an interlocutory one. It observed that the order of summoning the accused to a trial has serious consequences since he has to appear and face accusations and criminal proceedings even if the order was irregular or incorrect. It relied on Madhu Limaye’s intermediate order concept and V.C. Shukla’s liberal construction of interlocutory orders to hold the same.
In 2023, when the previous lawyer of the accused failed to ask certain questions during cross-examination that had to be asked to ensure a fair trial, the present legal counsel through Legal Services Authority had requested another cross-examination of prosecution witnesses on grounds of fair trial. While the special court of NDPS had rejected the application, the Calcutta High Court allowed the appeal and held that summoning witnesses was not an interlocutory order.
Similarly, in 2024, the Tripura High Court in Sri Arjun Debbarma v. State of Tripura held that summoning of witnesses was necessary in that particular case for proper adjudication and therefore, the petitioner had to be given an opportunity to do the same. It maintained that adducing evidence in support of the defence was a valuable right and denial of such right was denial of a fair trial.
These cases, as intended by Madhu Limaye and V.C Shukla, have laid great emphasis on the rights of the accused and fair trial principles to classify these issues as not being barred from revision, i.e., not purely interlocutory. Such classifications ensure that justice is not denied.
Conclusion
S. 397(2) of CrPC was introduced with the intent of minimising delays in criminal trials by preventing revisions of purely interlocutory orders. However, the true challenge posed by the provision (now S. 438(2) BNSS) is not merely due to the lack of a statutory definition, but rather from inconsistent judicial interpretations, especially concerning the concept of ‘intermediate order’ established by the Madhu Limaye case. While this landmark case established a balance between preventing delays and ensuring rights of the accused, subsequent misapplications and selective interpretations have led to judicial inconsistencies.
Recent rulings like Arjun Debbarma protect fair trials by allowing revisions of “intermediate” or “quasi-final” orders. However, this rights-based approach still carries a risk of judicial misinterpretation.
Therefore, even while BNSS did not attempt to provide a definition or any clarification of what interlocutory orders are, a right-sensitive interpretation of intermediate orders by courts will maintain the balance that the section intended to achieve, i.e., preventing delays while safeguarding rights of the accused.
Barbara Prem is a second-year student of the B.A. LL.B. (Hons.) course at the National Law School of India University (NLSIU), Bengaluru.
NUALS Law Journal 