(dissenting). If I could make myself believe that “because of the fog and the darkness Sloan [decedent] •assumed the duty of throwing the switch for the movements leading up to the accident,” I would have no difficulty in agreeing with the majority opinion in this case. The case turns upon this point of fact. Although of opinion that the evidence was insufficient to support a verdict for plaintiff, the trial court submitted the case to the jury, who found by special verdict that plaintiff’s decedent, Frank Sloan, and his two associate switchmen," Kessler and Hawkins, were each guilty of negligence directly contributing to the injury and death of Sloan, but that the negligence of Sloan was slighter. *651The jury also found that it w.as Hawkins’s duty to see that the switch in question was properly turned at the time and place in question, and that the decedent did not give the signals for the moving of the train immediately prior to the injury which resulted in his death. This verdict would, under the statute in force, have entitled plaintiff to judgment for the damages of $2,000 found. But the learned circuit judge changed the answers of the jury so as to say that the negligence of the deceased was greater, and so as to say that it was Sloan’s duty to see that the switch was properly turned, and upon the verdict thus changed gave judgment for defendant.
Sloan was foreman of the switching crew, which, excluding the engineer and fireman, consisted of Sloan, Kessler, and Hawkins. The only eye-witnesses were one Harry Gifford, an employee of defendant but not a member of Sloan’s crew, and Kessler and Hawkins. The injury occurred at 11:40 p. m. and the night was dark and foggy. A series of quite parallel switch tracks lay northerly of the main track and extended with it northeasterly and southwesterly. The witnesses refer to the first as “east” and to the second as “west.” That track next northerly of the main track is known as track 1; next north of this is track 2; next north of this lie tracks 3 and 4 in the order named. There are other tracks substantially parallel- with and north of track 4 and other tracks south of the main track, but they cut no figure in this case. It is seven and one-half feet from the north rail of track 1 to track 2, six feet between'tracks 2 and 3, and seven feet between tracks 3 and 4, all at a point opposite the switch shanty which is called in the evidence Camp 20. Southwesterly of Gamp 20 these tracks connect with switches. T*he distance from switch 1 northeasterly to C-amp 20 is 200 feet. Switch 1 connects track 4 with track 1. The distance from switch 1 northeasterly to switch 2 is ninety feet. Switch 2 connects track 4 with track 2. The distance northeasterly from switch 2 to switch 3 is eighty-nine feet. Switch 3 con*652nects track 4 with track 3. It will thus be seen that between tracks 1 and 4- are tracks 2 and 3, the southwesterly end of track 4 forming what is called in railroad parlance a lead for tracks 2 and 3; that with switch 1 open for track 4, cars would be shunted from track 1 onto track 4; moving along that track ninety feet they would encounter switch 2, which if opened for track 2 would shunt the cars onto track 2; but if switch 2 was not opened, moving along further northeasterly eighty-nine feet on track 4 they would come to switch 3, which if opened for track 3 would shunt the cars into track 3. Sloan was found lying dead directly south of the switch shanty, with his right leg and arm on the north rail of track 1, his lantern on the platform of the caboose, which had been backed down on track 1 northeasterly of where his body was. He had his train lists in his dead hand.
The train movements preceding the injury were -as follows: A switch train with thirteen or fourteen cars and a caboose was on track No. 2 and desirous of putting some of these cars but not the caboose on track 4. They desired to put the caboose on track No. 1. The caboose was the last or rear car. To do this they pulled southwest, passed over the switch connecting track No. 2 with track No. 4, over track No. 4 to switch No. 1, connecting track No. 4 and track No. 1, and far enough southwest on track No. 1 to clear switch No. 1, and then far enough northeast so as to back the caboose in on track No. 1 past switch No. 1 and uncouple it. It required three moves to do this: one movement to the southwest on tracks 2, 4, and 1 over switches 2 and 1, then one movement to the northeast on track No. 1, then uncoupling the caboose, then a “kickback” to the northeast so as to cause the caboose to clear the other cars which might come in on the switch track. Having got the caboose down on track No. 1 to a point about due south of Camp 20, they pulled southwesterly again along track No. 1 far enough to clear switch No. 1, and backed in on signals believing that switch *653No. 1 bad been thrown and that- they were backing in on track No. 4, bnt instead backing in directly on track No. 1, where they struck and killed Sloan.
It is very obvious that the neglect to throw switch No. 1 for the last backward or northeast movement and so shunt the cars onto 'track 4 was the negligent act which caused the injury in question. If we were to accept the words of the switchman Kessler without discrimination, the decision of the trial court might be sustained, because he said Sloan told him just before they pulled out over No. 2 switch that he, Sloan, would turn switch No. 1. But there is also other evidence, 'and therein lies the error in taking this ease from the jury. This evidence consists of the circumstance that Sloan was found dead on track No. 1 clutching his train sheets in his hand and with his lantern resting on the southwesterly platform of the caboose. Next, of the testimony of Harry Gifford, who sat in the door of Camp 20 looking toward Sloan and part of the time talking with him, and about seventy-five feet distant. He testifies:
“Sloan stayed there by the caboose when the train went west. He and I joshed back and forth, talking. He was looking over his list. His lantern was either on his arm or on the platform of the caboose. The train pulled west on number 1 track. Sloan gave no signals. This caboose was in front of the shanty. Sloan hollered to the rest of the crew to come down either 2 or 4 track, I couldn’t say which. After he called he stayed right by the caboose. He was standing there, and then I saw the cars coming back there on number 1 track. Then I saw him get killed. At the instant he got killed I put my hands to my ears. I thought he was going to holler.”
There are some contradictions and some discrepancy between this testimony and that which he gave at the inquest,' but nothing to take the question of his credibility from the jury. Gifford further testified that the first he saw of Hawkins the latter was walking northwest toward Camp 20 after *654the accident, coming from tbe southeast. This was not coming from the direction of track 4. Kessler, for. the defense,' testified that he was head brakeman and Hawkins rear brakeman; that Sloan told them to take the caboose out of track No. 2. They then went west, the engine facing east, and Sloan gave the signal to him to- stop when they pulled out from No: 2 west. Sloan then went west across No. 1 track— to throw the switch he supposed.
“If it had been a clear night it would have been my business to throw that switch, but he told me to stay with the engineer where he could see me and he would take care of the switches. I moved near enough to the engineer for him to see my signals. Sloan ought to have thrown that switch. Sloan gave me a signal to go 'ahead two or three car lengths. Eirst he gave a high ball. I gave them to the engineer. Erank was about seven or eight car lengths from me at that time. The engine then moved east. Sloan said he would throw the switch when he was right up to the engine coupling on No. 2. He then started to walk east on track No.
It will be observed that walking east on track'No. 2 was walking away from switch No. 1. It will also be observed that Mr. Kessler entirely omits the forward-and-back movement which they made after leaving track No. 2 with the caboose for the purpose of putting it on track No. 1, and the second back movement on track No. 1 for the purpose of kicking the caboose northeasterly to clear the switch. Now, if at the time they moved out of track No. 2 Sloan said he would throw the switch, he evidently did throw the switch, because after that they had deposited the caboose on track No. 1, and that could not have been done unless the switch was thrown, for track No. 1. So it is evident the testimony of this witness does not relate to the movement which resulted in the death of Sloan, but to a prior movement, and that he is concealing something. He continues:
“The last I saw of Sloan he gave me a car-length signal about four or five minutes before he was killed. The cars *655kept moving right along. We went one or two car lengths after Hawkins gave me the signal to stop. Hawkins came from the north. . . . After Sloan said he would throw the switch he walked east down No. 2 track. I was on the ground. At the time we pulled, across the switch. Sloan was nearer.”
From this it is apparent that both of these men were near the switch No. 2, both on the ground, and that Sloan proceeded east, which was away from switch No. 1, and that the other man, if he followed his train westwardly, passed switch No. 1. On cross-examination he explains that he did not see Sloan give signals, but saw the lantern signals and thinks Sloan gave them. That the person who gave him the highball signal to come ahead was at switch 1 where he went across. Tie testifies that he last saw Hawkins on track No. 2 before the accident when they pulled out from that track, and did not see him again until the accident. Hawkins was then on track No. 2, about ten car lengths from switch 2, and walking east. He reiterates that Sloan told him that he, Sloan, would throw switch 1 just before they putted out over switch 2. They were to come over No. 2 switch and over No. 1, and No. 1 switch was to be turned and Sloan turned it. This, I think, shows that he is not talking about the movement made at the time of the injury, but a prior movement. He says:
“We pulled west out of track No. 2 and backed the caboose on track No. 1. It would not clear, and then we shoved against it again and thought it cleared. Then he gave me the signal to go west over switch 1. I thought he would turn the switch so we would go dotun on lead Jf. Instead of this we went back on No. 1 track.”
As I understand this testimony, it does not contradict that of Gifford, and the statement of Sloan that he would turn the switch related to the first movement to move the caboose from track No. 2 onto track No. 1. That the switch was turned and this movement was made, and a second movement to kick *656the caboose back further so as to clear the switch, and then another southwesterly movement, and then the last and fatal movement, is beyond dispute.
Now it cannot be said that because Sloan, on the first movement, told Kessler he would throw switch No. 1 to make the connection necessary to place the caboose on track No. 1, that he would continue to throw switch No. 1 for any future movement. Having got the caboose at the .desired place he gives orders to his men to put the other cars on track No. 4 and applies himself to an inspection of his train sheets. These men then scatter. Kessler goes west, either riding or walking, passing over switch No. 1; Hawkins, according to his story, goes north to No. 4 track to open the knuckles on the cars there, and the train, after clearing No. 1 switch, and without knowing whether it has been turned or not, backs in again on No. 1 track upon Sloan while he is engaged in examining his train sheets. Kessler testifies to facts which would indicate that the signals for the last northeast movement, which resulted in Sloan’s death, were.given by Sloan. Gifford, who was watching Sloan at the time and talking with him, says Sloan gave no signals. The fact that Sloan stood right in the path of the train examining his train sheets by the light of his lantern is strong evidence to show he did not think the train was backing in on that track. If he gave the signals he must have understood that switch No. 1 was open for track No. 4. He knew he had not opened it. If he did not give the signals, as Gifford testifies, then Hawkins gave the signals and gave them recklessly,' without reference to whether switch No. 1 was turned for track No. 4 or track No. 1.
The case is not very large or very important, but after a man’s lips are closed in death I do not like to see his legal rights sworn away by the suspicious testimony of survivors interested in exonerating themselves. I think under the circumstances the evidence should be carefully scrutinized, and *657a careful scrutiny of tbis evidence convinces me tbat there were facts in controversy determinative of tbe case upon which a jury should have been allowed to pass. I think any experienced railroad switchman would unhesitatingly say that when a switching foreman, after having thrown a switch, walks away from it, and after the cars are thrown in to the required track and the foreman then gives another order to his men to put cars in on another track, such latter order includes passing beyond the switch so as to clear it and throwing the switch for that other track. I think that is what the evidence shows was done in this case.
I cannot assent to the correctness of the following in the opinion of the court:
“All reasonable doubts arising from the record itself, and those arising from appreciably contemplating the peculiarly advantageous position of the trial judge, must be overcome, in such a case as this, in order to successfully challenge the initial decision.”
As I understand this, it means to say that the decision of the trial court changing the answer of the jury to one or more questions of a special verdict must be affirmed, unless the contrary appear heyond all reasonable doubts growing out of or suggested by the printed evidence or by considering the advantage which the circuit judge had in seeing and hearing the witnesses and observing their demeanor and manner of testifying. I do not think such weight can he given to the decision of the court below upon what is in legal effect a demurrer to evidence, without invading the constitutional right of the suitor to a trial by jury.
In Klein v. Valerius, 87 Wis. 54, 57 N. W. 1112, a statute was held unconstitutional which attempted to make it the duty of this court to review all questions of law or fact presented by the record upon appeal or writ of error and to examine and review the evidence when the same is preserved by a bill of exceptions and give judgment upon the right of the *658case, regardless of the decisions upon questions of fact or law made by the court below. It was suggested that the determination of questions of fact must remain as at the time of the adoption of the constitution, citing Callanan v. Judd, 23 Wis. 343. Sec. 5, art. I, Const., secures the right of trial by jury in civil cases. Bennett v. State, 57 Wis. 69, 14 N. W. 912. Where the question of negligence is of a mixed character, one of law and fact, it must be left to the jury. Patten v. C. & N. W. R. Co. 32 Wis. 524. Mixed questions of law and fact are to be submitted to the jury under proper directions. Bass v. C. & N. W. R. Co. 36 Wis. 450. Where conflicting and indecisive evidence is given upon questions of fact it is for the jury to determine its force and effect. Saunderson v. Lace, 2 Pin. 257. Where there is evidence for each party upon a material point, though positive upon the one side and indefinite upon the other, the weight of such evidence is a question for the jury. Calder v. Crowley, 74 Wis. 157, 42 N. W. 266. This is also the rule where the solution of the question depends upon circumstantial evidence. Agnew v. Farmers’ Mut. P. F. Ins. Co. 95 Wis. 445, 70 N. W. 554. The mere fact (as shown by the record in this case) that the trial judge believes that if a verdict should be rendered for the plaintiff in a personal injury case it ought to be set aside as against the weight of evidence, does not justify him in withdrawing the case from the jury. O’Brien v. C. & N. W. R. Co. 92 Wis. 340, 66 N. W. 363.
Many such cases might be cited, and in the early history of this court this command of the constitution lay with serious and heavy control upon the judges of this court. As a logical consequence of this view there obtained also the rule that a motion for a nonsuit presented a question of law, the facts being admitted with all their proper effect in law (Hunter v. Warner, 1 Wis. 141), and that a nonsuit for insufficiency of evidence could be granted only where there was no evidence which upon the most favorable construction it would *659bear and no inferences of fact capable of being logically drawn therefrom which would justify the jury in finding for the plaintiff. Douglass v. Garrett, 5 Wis. 85; Johnston v. Hamburger, 13 Wis. 175; Jarvis v. Hamilton, 19 Wis. 187; Sutton v. Wauwatosa, 29 Wis. 21; Gower v. C., M. & St. P. R. Co. 45 Wis. 182; Schomer v. Hekla F. Ins. Co. 50 Wis. 575, 7 N. W. 544; Imhoff v. C. & M. R. Co. 22 Wis. 681; Houfe v. Fulton, 29 Wis. 296.
Many other cases might be cited. The first invasion of these rules is found, I think, in Powell v. Ashland I. & S. Co. 98 Wis. 35, 38, 73 N. W. 573. While professing to follow the prevailing rule of law above referred to, several qualifications are added. The legal effect of evidence is to be measured by a new standard, that is, it must, in order to be taken from the jury, be so clear and conclusive as not to admit reasonably of any opposing inferences in unbiased and unprejudiced minds. There is an element of confusion in the rule thus stated and an argumentum ad hominem which enables the court applying the rule to easily dispose of cases on the ground that those differing did not possess the unbiased and unprejudiced mind. But this apparently easy avenue of escape for one who had the ultimate disposition of a controverted question failed or reduced its,elf to- absurdity when the tribunal of last resort divided upon the question, as in Agen v. Metropolitan L. Ins. Co. 105 Wis. 217, 224, 80 N. W. 1020; Sparks v. Wis. Cent. R. Co. 139 Wis. 108, 120 N. W. 858; Zabawa v. Oberbeck Bros. Mfg. Co. 146 Wis. 621, 131 N. W. 826; and other cases. A decision under the rule stated does not decide the case, but merely decides our moral and mental superiority over those who differ with us as to the probative effect or proper inferences to be drawn from facts.
When one person or tribunal decides upon given evidence that it contains nothing either by way of direct proof or logical inference to support a stated conclusion, and another undertakes to review and reverse that decision, the latter *660must point out tbe error and. show tbe items of direct proof or logical inference wbicb have been overlooked or undervalued by tbe first decision. But no such thing is necessary if tbe rule is so modified as to permit tbe reviewer to bolster up bis own argument by determining tbat be was intelligent, unbiased, and unprejudiced, while all wbo differed with him lacked those high qualities. But the decision in Powell v. Ashland I. & S. Co., supra, did not stop here. It undertakes to add to the existing rule, and declares that, while we must give the evidence and the logical inference therefrom this effect most favorable to the person nonsuited and consequently against tbe decision, we must give the decision such weight that it cannot be disturbed unless against the clear preponderance of the evidence. This is impossible, contradictory, and most confusing. No one can reconcile these two rules. The preponderance of evidence has been defined, and it has been often said that that evidence preponderates which is the most credible and convincing. Parker v. Hull, 71 Wis. 368, 31 N. W. 351. I do not attach great weight to judicial definitions except as they affect the particular case. It is not a judicial function to define the meaning of a word as applied to all cases, and judicial definitions from this viewpoint are more frequently wrong than right, except when they perform the idle office of copying from the current dictionaries. But tbe words “preponderance of evidence” naturally and ordinarily refer to a question of fact, i. e. the difference in convincing power between two opposed probative tendencies. If there could be any doubt on this, it must be dissipated by the further language of the court in Powell v. Ashland I. & S. Co., supra:
“Tbe opportunities wbicb a trial court has for determining sucb questions are superior to those possessed by the appellate court,' and on that ground are entitled to considerable weight, and, where the question is doubtful, to controlling weight.”
*661This language is applicable only to questions of fact arising in the court below, and the expression, “where the question is doubtful to controlling weight,” is in direct contradiction of that part of the rule as stated in the same case, which declares that “unless the proof of contributory negligence is so clear and decisive as to leave no room for unbiased and impartial minds to come to any other conclusion, . . . the proper inference to be drawn must be determined by the jury.” Upon this background of confusion in thought and language this court has attempted for some years to sketch a plan of procedure to guide professional men, inferior courts, and itself. It has succeeded in proving under this rule, in Agen v. Metropolitan L. Ins. Co. 105 Wis. 217, 80 N. W. 1020, that either three of the judges of this court, or else two of the judges of this court plus the circuit judge and the jury, did not possess unbiased or unprejudiced minds. Very many conflicting and confused statements of this alleged rule are thereafter to be found in the reports. In Turtenwald v. Wis. Lakes I. & C. Co. 121 Wis. 65, 98 N. W. 948, where the ruling of the circuit court directing a verdict for defendant was reversed, the rule was softened down as follows:
“Ordinarily a trial judge in deciding whether the evidence in any reasonable view of it will support a verdict in favor of the plaintiff has some advantages over an appellate court, and that such circumstance, in doubtful cases, is entitled to considerable weight upon appeal.”
There will also be found the aggressive and militant form of the rule, as in Clemons v. C., St. P., M. & O. R. Co. 137 Wis. 387, 391, 119 N. W. 102, where it is said:
“A conclusion of a trial court, respecting sufficiency of evidence as to any fact in issue to present a jury question, should not be disturbed unless it appears from the record to be clearly wrong, giving due weight to the superior advantages which such court has for discovering the truth."
*662This alleged rule bas been stated in many ways,, differing not only in form but in substance, and sometimes without distinction of the question presented. When the trial court bas decided that there is evidence sufficient to take the case or the question to the jury and the jury bas decided affirmatively or negatively on that question, and it is sought to review the said ruling of the trial court in this court, the double-beaded rule declared in Powell v. Ashland I. & S. Co. 98 Wis. 35, 73 N. W. 573, could be logically applied, leaving out the personal element; because in that case the rule which entitles one to a jury trial upon questions of fact and entitles him to have the evidence submitted to the jury, if, viewed in its aspect most favorable to him, there is any basis for a jury finding, and the rule that the decision of the circuit court is entitled to weight, would coincide and tend in the same direction; but where the question for review in this court is a ruling of the trial court taking the case away from the jury on account of insufficiency of evidence, then the double rule in question is contradictory and inconsistent. Add to this confusion that several justices of this court seem to have a different way of stating the rule, and we have a deplorable situation. See Zabawa v. Oberbeck Bros. Mfg. Co. 146 Wis. 621, 131 N. W. 826, dissenting opinion; Collier v. Salem, 146 Wis. 106, 130 N. W. 877; Lines v. Milwaukee, 147 Wis. 546, 133 N. W. 592; Collins v. Janesville, 117 Wis. 415, 423, 425, 94 N. W. 309; McCune v. Badger, 126 Wis. 186, 189, 105 N. W. 667; Fleming v. Northern T. P. Mill, 135 Wis. 157, 114 N. W. 841. “Considerable reliance,” Lind v. Uniform S. & P. Co. 140 Wis. 183, 188, 120 N. W. 839; Fitzpatrick v. Lake Superior T. &. T. R. Co. 142 Wis. 65, 70, 124 N. W. 1054.
But the further “reasonable doubt” addition to this rule applied in the present case goes beyond anything heretofore approved by this court, although there is a suggestion of it in a dissenting opinion by Mr. Justice Maeshall in Sparks *663v. Wis. Cent. R. Co. 139 Wis. 108, 120 N. W. 858. The confusion that has crept into the courts on this subject is well displayed in the decision in Kroger v. Cumberland F. P. Co. 145 Wis. 433, 130 N. W. 513, and perhaps it was well to say, as was said in Meyst v. Frederickson, 146 Wis. 85, 130 N. W. 960: “This rule is not an arbitrary one, but applies with varying strength according to circumstances.” This is equivalent to saying this rule is not a rule. True, there is no rule without its exceptions, bitt an exception is not a question of varying strength. The instant case must have been one for the application of the rule with great strength. When the jury and the circuit judge both pass on the amount of the damages suffered in a personal injury case, this rule in favor of the decision of the circuit court seems to lose its force altogether, and this court often undertakes to disagree with the court below, yet there must in nearly all such cases be a reasonable doubt with reference to the amount of damages actually sustained. Olwell v. Skobis, 126 Wis. 308, 105 N. W. 777; Koepp v. Nat. E. &. S. Co., ante, p. 302, 139 N. W. 179.
If this were a rule of property or a question subject to the maxim sta/i'e decisis, it might be considered rash at this late day to suggest a correction,- and in such case we would be obliged to call upon the legislature for relief. But that branch of the government is hardly able to give us relief from the infirmity displayed by the alleged rules. Neither can it be claimed that a decision by the supreme court that confusion is clear will have the desired effect. The situation seems to call for a'new application of the “rule of reason” to a much beclouded situation.
I am authorized to say Mr. Justice KerwiN concurs in this dissent.